WO2001011894A2 - Methods and apparatuses for encoding and displaying stereograms - Google Patents

Methods and apparatuses for encoding and displaying stereograms Download PDF

Info

Publication number
WO2001011894A2
WO2001011894A2 PCT/DK2000/000448 DK0000448W WO0111894A2 WO 2001011894 A2 WO2001011894 A2 WO 2001011894A2 DK 0000448 W DK0000448 W DK 0000448W WO 0111894 A2 WO0111894 A2 WO 0111894A2
Authority
WO
WIPO (PCT)
Prior art keywords
image
colour
stereogram
component
filter
Prior art date
Application number
PCT/DK2000/000448
Other languages
French (fr)
Other versions
WO2001011894A3 (en
WO2001011894A8 (en
Inventor
Per Skafte Hansen
Original Assignee
Per Skafte Hansen
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Per Skafte Hansen filed Critical Per Skafte Hansen
Priority to AU64275/00A priority Critical patent/AU6427500A/en
Priority to EP00951279A priority patent/EP1245121A2/en
Publication of WO2001011894A2 publication Critical patent/WO2001011894A2/en
Publication of WO2001011894A3 publication Critical patent/WO2001011894A3/en
Publication of WO2001011894A8 publication Critical patent/WO2001011894A8/en

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/30Image reproducers
    • H04N13/324Colour aspects
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/20Image signal generators
    • H04N13/204Image signal generators using stereoscopic image cameras
    • H04N13/239Image signal generators using stereoscopic image cameras using two 2D image sensors having a relative position equal to or related to the interocular distance
    • GPHYSICS
    • G02OPTICS
    • G02BOPTICAL ELEMENTS, SYSTEMS OR APPARATUS
    • G02B30/00Optical systems or apparatus for producing three-dimensional [3D] effects, e.g. stereoscopic images
    • G02B30/20Optical systems or apparatus for producing three-dimensional [3D] effects, e.g. stereoscopic images by providing first and second parallax images to an observer's left and right eyes
    • G02B30/22Optical systems or apparatus for producing three-dimensional [3D] effects, e.g. stereoscopic images by providing first and second parallax images to an observer's left and right eyes of the stereoscopic type
    • G02B30/23Optical systems or apparatus for producing three-dimensional [3D] effects, e.g. stereoscopic images by providing first and second parallax images to an observer's left and right eyes of the stereoscopic type using wavelength separation, e.g. using anaglyph techniques
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/20Image signal generators
    • H04N13/204Image signal generators using stereoscopic image cameras
    • H04N13/243Image signal generators using stereoscopic image cameras using three or more 2D image sensors
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/30Image reproducers
    • H04N13/332Displays for viewing with the aid of special glasses or head-mounted displays [HMD]
    • H04N13/334Displays for viewing with the aid of special glasses or head-mounted displays [HMD] using spectral multiplexing
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/30Image reproducers
    • H04N13/363Image reproducers using image projection screens
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/10Processing, recording or transmission of stereoscopic or multi-view image signals
    • H04N13/106Processing image signals
    • H04N13/15Processing image signals for colour aspects of image signals
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/10Processing, recording or transmission of stereoscopic or multi-view image signals
    • H04N13/189Recording image signals; Reproducing recorded image signals
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/20Image signal generators
    • H04N13/257Colour aspects
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/20Image signal generators
    • H04N13/286Image signal generators having separate monoscopic and stereoscopic modes
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/30Image reproducers
    • H04N13/332Displays for viewing with the aid of special glasses or head-mounted displays [HMD]
    • H04N13/337Displays for viewing with the aid of special glasses or head-mounted displays [HMD] using polarisation multiplexing
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/30Image reproducers
    • H04N13/332Displays for viewing with the aid of special glasses or head-mounted displays [HMD]
    • H04N13/341Displays for viewing with the aid of special glasses or head-mounted displays [HMD] using temporal multiplexing
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/30Image reproducers
    • H04N13/361Reproducing mixed stereoscopic images; Reproducing mixed monoscopic and stereoscopic images, e.g. a stereoscopic image overlay window on a monoscopic image background
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/30Image reproducers
    • H04N13/365Image reproducers using digital micromirror devices [DMD]
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/30Image reproducers
    • H04N13/398Synchronisation thereof; Control thereof
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/50Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
    • H04N19/597Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding specially adapted for multi-view video sequence encoding

Definitions

  • the present invention relates to the encoding, or recording, and subsequent displaying and viewing of a stereogram as one colour image for viewing through a pair of filters, the filters transmitting light m the visible spectral range, each with a different transmission profile.
  • Colour encoded stereograms are conventionally known as anaglyphs, although this term m the present context will be reserved for a sub-class of colour-encoded stereograms. This subclass of images contains the vast ma ority of images made with the prior art and will be more precisely defined shortly.
  • the viewing filters must be separating. This word is here chosen to mean that there must be a first range of the visible spectrum transmitted by a first filter and excluded by a second; and a second range of the visible spectrum, not overlapping the first range and transmitted by the second filter and excluded by the first.
  • luminance level dm- age will be used, whenever the intended meaning is that only one channel or colour variable is represented (m the image) .
  • the visual appearance of a luminance level image may or may not be gray. If it is not gray, only one nameable colour (i.e. one hue) is present m it, m varying intensities.
  • the fil- ters will m the present context be called "disjoint" (although the word may have other connotations). Also for convenience, the full range of the visible light can be divided into three: visible light of wavelengths less than 500 nanometers, light m the range from 500 nanome- ters to 600 nanometers, and visible light having wavelengths greater than 600 nanometers.
  • a filter pair for anaglyph viewing can be termed a (1,1) -pair, if each filter transmits light from only one of these three ranges, and a (1,2) -pair if one filter transmits light from only one range, the other from the remaining two.
  • a red-green filter pair is then denoted as a ([1 0 0], [0 1 0])-pa ⁇ r, a red- cyan filter pair as a ([1 0 0], [0 1 l])-pa ⁇ r etc.
  • cyan is complementary to red, the word "cyan” being here used to denote an ideal mixture of green and blue.
  • a red-cyan filter pair is disjoint
  • the numbers 1 and 0 do not literally mean “all” and “none”, respectively, as no physical filter transmits all light m one spectral range and no light m another. Also, where the need arises to balance filters for complementarity m the narrower sense mentioned above, exact numbers would yield ratios appreciable different from 1:1.
  • a stereogram will be said to consist of two part-images, a left and a right. Terms such as “half images” are some- times seen, but it must be kept m mind that the two constituents of a (colour) stereogram are full two- dimensional (colour) images m themselves.
  • the aim of the art is to present a stereogram m such a fashion as to allow the eye-brain system to per- form fusion of colours as well as stereoscopic interpretation, and to eliminate or at least reduce rivalry where it should not appear.
  • the lustre of smooth fabrics, the gloss of polished metals and the glimmer of gems are all perceptive phenomena, produced by rivalry. These, of course, should not be suppressed.
  • a special form of stereoscopic rivalry can only be reduced, not removed: when a specific colour is transmitted by only one of the viewing filters, it may be perceived as having the right or nearly the right hue (the nameable quality, such as "pink”, “lavender”, “moss green”), but being materially different from other colours in the image.
  • the nameable quality such as "pink”, “lavender”, “moss green”
  • reflection prints it may e.g. seem translucent, on emissive screens it may e.g. take on a perceived luminance m excess of what the display can otherwise deliver.
  • Stereopsis is usually achieved, except when the original stereogram relies on, say, texture perspective as a support of parallactic differences, and this texture perspective happens to suffer from the colour loss.
  • Large areas of a pure colour can loose their stereoscopic ef ⁇ fect, if the colour is only transmitted by one of the viewing filters.
  • Rivalry, over and above the "sheen" is typically seen where strong contrasts exist m the stereogram, or where the surrounds of the display are not sufficiently dimmed, and may m some individuals lead to discomfort (nausea, headaches etc.), sometimes ascribed to "colour bombardment". Diplopia can occur, over and above what is caused by transgression of stereoscopic parameter limits, when the recording medium, the display or the filters, or any combination of these, fail to separate the part-images of the stereogram.
  • Patent US 5491646 diplopia caused by insufficient separation is subjected to the countermeasure of subtracting a scaled copy of the colour values extracted from one part-image from those extracted from the other "so that said first image is entirely, or almost entirely removed from said second image", a digital counterpart of the masking technique used m printing.
  • the prescriptions m Patent US 5491646 otherwise address the use of (1,2)- fliters .
  • Patent US 4217602 use is made, not of two part- images, but three.
  • the purpose is to allow non- stereoscopic, full-colour viewing, the stereoscopic ef- feet being obtained by the use of ( 1 , 1 ) -filters and thus giving a "monochrome" stereo image m the sense described above and discarding the third image.
  • the tnree lenses or lens systems of Patent US 4217602 are identical.
  • Patent US 4620770 use is made of two outline images ana a colour filling, rather than two definite images.
  • the method as described pertains to hand-drawn a anaglyphs, the viewing filters are referred to simply as "3- D glasses" and appear to be intended as (1,2) -pairs, and no procedure is given for the production of specific col- ours .
  • One of the filters m the pair still transmits from only one of the tnree colour ranges, though; there is a small amount of "ghost imaging"; and at least m the blue-amber combination here given as an example, the blue filter transmits only a small portion of the actual light.
  • Valyus (op. cit.) p. 109, describes the use by L. Lumiere of a blue-yellow filter pair which m fact, according to the few spectral data cited, must have been a violet- yellow ([1 0 1/2], [0 1/2 l/2])-pa ⁇ r. Obviously, L. Lumiere can not have attempted to display full-colour lm- ages - m fact, Valyus writes: "One of his [L. Lumiere's] images was coloured yellow, the other blue", suggesting "monochrome” anaglyphs - and there is no explicit record of further attempts with filters of this kind.
  • a “channel” is here a single valued function, such as a luminance level function, defined on the image area.
  • the function may take values m a circular domain, as with the hue, cp . the expression “hue circle”) .
  • a further two channels are required, typically representing the hue and saturation, respectively, of each scene point.
  • a minimum of four colour channels is required to produce a full- colour stereogram.
  • m a viewing filter pair designed as prescribed by the invention, both of the filters must transmit light m two of the ranges red, green or blue; to avoid double imaging and allow the use of optimisation and colour management, recording and encodings are no longer pomt-to- point; and display, even of existing stereograms, is colour-balanced, as far as the pertinent choice of media allows .
  • the present invention offers m its different aspects
  • a filter pair as prescribed by the invention must then be a "(3/2, 3/2) -pair" m the sense that the filters must be separating with respect to the carrier colour ranges; but both filters must transmit the mediant colour range.
  • This latter can take the form of a joint transmission m the entire mediant colour range; or each filter can transmit a part of this range.
  • Transmission m the mediant colour range may involve some reduction, relative to transmission m the pertinent carrier colour range.
  • the phrase "colour range” will be replaced by the simpler "colour”.
  • three separate images can then be recognized: one m each of the carrier colours and one in the mediant colour.
  • the same filters can be used as barrier fil- ters m a display device. If this device is a conventional stereoscopic projector, the filters can thus take the place of polarizing barrier filters. In a digital projector, the filters can again take the place of polar- lzing filters, but there is the further advantage that m the left part-image, no constituent corresponding to the carrier colour of the right part-image need ever be formed (or it can be suppressed); and vice versa.
  • Such a digital projector therefore, m a sense, projects four colour channels: the carrier colour associated with the left part-image and transmitted by the left-image filter; the mediant colour associated with the left part-image, but projected only m the sub-range transmitted by the left-image filter; the mediant colour associated with the right part-image, but projected only m the sub-range transmitted by the right-image filter; and the carrier colour associated with the right part-image and transmitted by the right-image filter.
  • the requirement of four perceptual colour channels is met directly and can be fully utilized by an optimising encoding of the stereogram, while m the remaining encodings described below, primarily aimed at conventional three- channel displays, the impression of a four-channel display is created by colour management.
  • the two carrier colour images considered as images m themselves, must, at least to some extent, compensate for any loss of sharpness caused by this blurring of the mediant colour image.
  • the amplitudes of the left carrier colour image will be enhanced, wherever the amplitude of the mediant colour image is smaller than that of the mediant colour component of the original left part-image; and it will be dampened, wherever the amplitude of the mediant colour image is larger than that of the mediant colour component of the original right part- image.
  • the right carrier colour image mutatis mutandis
  • an optimisation process will, when it is operating on a so-called "cost function" that also takes into account the final per- ceived colours, select those that cause the smallest changes (and hence, if possible: no changes) m the fused colours, compared with the fused colours perceived m the original stereogram, when this is stereoscopically viewed.
  • the encoding must replace some amount of one or both carrier colour with mediant colour, and vice versa, and this replacement can not be performed on a pomt-by-point basis, as this would create visible dis- continuities that would give rise to both diplopia and abrupt colour casts.
  • the simplest instantiation of the encoding is obtained by first photographically recording, along three parallel lines of sight, three (luminance level) images, the middle one having a restricted depth of field m the sense of being visually sharp only m the distance range m- tended for zero parallax m the final stereogram, then encoding the outermost two m the respective carrier colours and the middle one in the mediant colour, and finally mounting the three m register, i.e. with the zero parallax motif parts coalescent and homologous points horizontally aligned, to form a full-colour image.
  • colour separation filtered images can be recorded.
  • a re- cording process of this kind can be accomplished by an optical apparatus, but it should be noted that such an apparatus performs no amplitude enhancements of the carrier colour constituents.
  • the process can also be accomplished by an electronic filter, forming part of e.g. an electronic camera or image recorder.
  • Digital image recorders can avoid the double imaging caused by spectral separation failures; and an electronic filter can perform enhancements of the carrier colour constituents and indeed other cross-constituent operations, as they are described m the following.
  • the description of the image encoding as prescribed by the invention can proceed m steps, each resulting instantiation being a variation over or an improvement of this theme:
  • the middle image need not be directly recorded, but can be computed as an average of the mediant colour "planes" of the two outer images, each of which must then be recorded in at least two colours, the mediant colour and the relevant carrier colour.
  • the averaging can be simple (half the sum), weighted using pre-determmed weights, or adap- tively weighted, taking into account the fact that areas which are low in one of the carrier colours provide less room for subsequent colour manipulations.
  • the optical blurring can be simulated by a smoothing process.
  • the smoothing operator can be a simple "sliding strip” technique, one or two-dimensional, or a genuine approximation to the point spread function of a lens with finite (as opposed to point-shaped) aperture.
  • the local strength of the smoothing can vary according to any measure of local parallactic separation. Assuming that the stereo registration has been decided, this can be put m simpler terms: if there is very little difference between the mediant colour contents (or, alternatively, the gray scale contents) of the left and right images m and around a given position, the mediant colour should be only gently smoothed, in and around that position.
  • Such a replacement may also be used to remedy double image phenomena: if, owing to the spectral imperfections of recording media, display media or filters, an "echo" of the left part-image is visible to the right eye or vice versa, the one carrier colour image may be modified by subtraction of a scaled version of the other carrier colour image. (Lest it may seem that this procedure leads to an infinite regress, it should be noted that the problem can be stated as two simultaneous equations m two unknowns; that the repeated procedure, if m fact performed, has the character of a "relaxation solution" of this problem; and that the smallness of the scaling factors implies a rapid convergence) .
  • a correspondence map of the stereogram is a representation, typically m the form of a gray scale image, of the separation at any location, of the point at that location m one of the part-images and its homologous point m the other part-image.
  • Homologous points are images of one and the same scene point. Their separation is geometrically related to the depth of the scene point. Algorithms for solving the correspondence problem exist m great number m the arts of photogram- metry and machine vision.
  • the stereogram can be artificially de- saturated with little loss of image quality: if the saturation of colours of distant areas is reduced, they will be more easily encoded without causing artefacts of dis- continuities m, or mis-coloration of, areas closer to the foreground. As with the adaptive averaging, this is again a question of providing room for colour re ⁇ distribution .
  • any given image area serves two purposes simultaneously: it is the representation of an area m the original left part-image - and a representation of a different area m the original right part-image.
  • a colour-encoded stereogram considered as seen by the left eye
  • a homologous area can be found.
  • the stereogram is assumed to be an endo- stereogram (all scene points are perceived as lying behind the image plane), so the homologous area lies to the right of the given area.
  • a new homologous area can be found further to the right, etc.
  • a model for what is actually perceived m a colour encoded stereogram must represent such chains of homologous areas, typically m the form of lists of selected representative homologous point pairs and their surroundings. Colour fusion will be between the colours seen m an area and its homologous counterpart. The areas must be found m (and referred back to) the original stereogram.
  • a norm or metric on a device-independent col ⁇ our space an algorithm to solve the correspondence problem; a model for colour fusion; spectral measurements and models to convert device and image colour values into de- vice-independent values; an algorithm for numerical opti ⁇ misation; a viewing filter pair chosen as prescribed by the invention; and a stereogram - the image encoding pre ⁇ scribed by the invention: solves the correspondence problem for the original stereogram, representing the solution as a list or lists of homologous areas; - computes the perceived fused colours m these areas, according to the fusion model; sets up the optimisation problem of encoding the stereogram m such a fashion as to express that what is actually perceived (as fused colours m the same areas) m the encoded stereogram, when this is observed through the viewing filters, has the smallest deviation m terms of the norm or metric on the colour space from the fused colours m the original stereogram;
  • display of the recorded or encoded images can be made using any conventional (colour) display medium, including the media of prints or colour photos. If circumstances allow, though, use can advantageously be made of special stereoscopic projectors (designed as prescribed by the invention) accommodating coloured filters, which must then be functionally matched by the viewing filters.
  • the transmittance over a range shall mean the area under the transmission rate function graph divided by the width the range; and a filter will be said to "transmit a range", if its transmittance over that range is larger than 0.06 (6 C ), and to "exclude a range” if its transmittance over that range is less than 0.06. (This value corresponds to a damping of four photographic "stops” and marks a practical limit at which e.g. the presence of "ghost images” begins to cause diplopia) . Two filters will be said to have a "significant common transmission" of a range, if the common area of their transmission rate functions restricted to that range is no less than 10 of either area.
  • a "correspondence map” is defined elsewhere m the text and it remains only to note, that it is understood that homologous point pairs are at least approximately brought to lie on parallel lines ("horizontal lines"), and that such a map may take on negative values (negative parallaxes , if part of the scene depicted lies front of the image plane used in the depiction. If the left and right part-images of an alleged stereogram are actually identical, the scene depicted is a flat object, parallel to the image plane, and its correspondence map is constant, possibly everywhere zero.
  • an image can be considered a two-dimensional signal, and concepts such as discrete Fourier transforms applied to it. Using such a transform one may define frequency ranges and their energy contents. Reference is made to the literature on the subject. Only the idea of “smoothing" an image by removing or reducing some of its “energy” m the “high frequency range” is needed m the following. Such a description is valid, even if the actual smoothing is carried out by other means.
  • Figure 1 is a diagram showing the signal routes and operational units of an apparatus that performs the encoding of a stereogram as prescribed by the invention
  • figure 2 is a diagram showing the functional elements an optical apparatus that performs the encoding of a stereogram as prescribed by the invention
  • figure 3 is a diagram showing the signal routes and operational units of an apparatus that performs the encoding of a stereogram as prescribed by the invention
  • figures 4 to 6 show idealized transmission rates of viewing filter pairs as prescribed by the invention
  • FIGS 7 to 9 show transmission rates of physical view- mg filter pairs designed as prescribed by the invention.
  • figures 10 to 12 show idealized transmission rates of filter pairs for use, as viewing filters, and, m the projection apparatuses prescribed by the invention, si- multaneously as barrier filters;
  • figure 13 is a diagram showing the application of a filter pair with spectral properties derived from the ideal transmission rates of one of figures 10 to 12 as barrier filters a conventional stereo projector;
  • figure 14 shows one light ray path and a possible positioning of one of the filters of figure 10 to 12, jointly with a colour separation filter, m a projector as pre- scribed by the invention.
  • a stereogram is given as two part-images digital form, represented by their "RGB" value arrays (as is customary m e.g. computer graphics).
  • RGB RGB value arrays
  • the left carrier colour is exemplified by B (blue)
  • the right carrier colour by G green
  • the mediant colour by R red
  • the R components of the left and right images are then halved, added up and subjected to a smoothing m the form of a "sliding strip” averaging with a weight function m the form of a "Gaussian bell" (both terms defined and described detail m standard refer- ences on numerical analysis and image analysis) .
  • the encoded image is obtained as the combination of this averaged R component with the B component from the left image and the G component of the right image.
  • the simplicity of this embodiment is evident, but it is only truly applica- ble to stereograms with low contrasts and restricted depth .
  • an approximate correspondence map is first obtained, for example by simply computing the point-wise differences m luminance levels everywhere over the two part-images of the stereogram (this is crude, but remarkably effective for its present use) .
  • An adaptive averaging of the two R- components of the part-images is then computed: First, a point-wise sum is formed of the minimum of the two R- values with a weighted sum of the two differences between the minimum and the actual values (thus, m any given point, one or the other of the contributions is zero) .
  • the weights are so chosen, relative to the luminances of re ⁇ , green and blue the ⁇ isplay medium and to a selec- tion of test images, so that colour casts m the test images are found acceptable by visual inspection.
  • the choice of weights is thus empirical (psycho- physiological ) .
  • the result is smoothed, as described with reference to the first embodiment, with the smoothing strength now coupled to the values of the correspondence map.
  • the coupling is again determined by visual inspection, m order to avoid diplopia.
  • the encoded image is obtained as the combination of this averaged R component with the B component from the left image and the G component of the right image.
  • encoding proceeds as m the second embodiment, but the B and G components are further corrected to eliminate or nearly eliminate luminance variations ("ghost images") caused by the failure of the recording medium or of the viewing filters relative to the display medium or both to preserve ideal spectral separation.
  • G images luminance variations
  • a genuine, if not exact, correspondence map is first obtained (by any method of choice) and the saturation of colours m the part-images of the stereogram reduced, m- creas gly with increasing depth, colours of the nearest points remaining unreduced.
  • Encoding proceeds as m the third embodiment, except that the degree of smoothing is now more precisely controllable, a more precise corre ⁇ spondence map being available.
  • an exact correspondence map can be generated m the course of the rendering process, either directly by the expedient of monitoring the projection of scene points, or indirectly from the so-called depth buffers of the part-images.
  • - a device-independent colour space with a metric (colour distance) defined on it (several such are described colour science) ; - an extension of the metric, such as an integration or weighted summation of point-wise values of the metric, to an error measure allowing comparison of colours of fused homologous areas, as these are seen m the original stereogram, and m the encoding, respectively; - a model for colour fusion (with additive mixture as the very simplest); and
  • the encoding is then found as the solution to the optimisation problem of reducing the error, as determined by the error measure, of the fused colours seen m the encoded image, compared with the fused colours m the stereogram, as these are determined according to the colour fusion model, acting over homologous areas m the chains partitioning the stereogram.
  • the form of apparatuses aiding the recording of stereograms m the form of apparatuses aiding the recording of stereograms :
  • a sixth preferred embodiment of the invention takes the form of an electronic filter converting three incoming full-colour (or: a two- colour, a one-colour and a two-colour) image signals to one outgoing full-colour image signal.
  • This embodiment comprises, as shown m figure 1:
  • - a component, 101 for separating the partial signal representing the first carrier colour and the mediant colour from the first image signal
  • - a component, 102 for separating the partial signal representing the mediant colour from the second image signal
  • a component, 130 for the combination of a modified first carrier colour image signal, a smoothed second mediant colour image signal (optionally modified) and a modified third carrier colour image signal, modifications being controlled by the outcome of the comparisons m the component 120.
  • Through 150 runs the signal representing the second mediant colour image.
  • Through 160 runs the signal representing the smoothed second mediant colour image.
  • Through 170 runs the (optionally modified) smoothed mediant colour image signal and signals representing the outcome of the com- pa ⁇ sons made m 120.
  • Through 180 runs the signal repre ⁇ senting the final, encoded image.
  • a seventh preferred embodiment of the invention takes the form of an optical adaptor converting three visual image signals to one outgoing visual image signal.
  • the embodiment comprises, as shown m figure 2 :
  • - optical colour separation filters 201, 202 and 203 can for instance be the filters most frequently used other photographic colour separation work, Kodak Wratten 25 (red) , Kodak Wratten 58 (green) and Kodak Wratten 47B (blue) m some order. (The spectral properties of these filters are described m Kodak's technical references, and also elsewhere) ; - lenses or lens systems, 210, 211 and 212, with 211 having a reduced depth of field relative to 210 and 212 (which are identical);
  • - prismatic mirrors 230 allowing passage of an lmage- forming light ray (bundle) from 211 and reflecting an image-forming light ray (bundle) from each of 220 and 221 such a fashion as to produce a combined image.
  • the device would be of fixed stereo- scopic registration.
  • a coupling (not shown separately in the drawing) between the lens system of the camera, on which the adaptor is mounted, and e.g. the lenses or lens systems 210 and 212 can be added to allow variable stereoscopic registration.
  • An eighth preferred embodiment of the invention takes the form of an elec- tronic filter to be mounted m a three-lens camera, m which colour separation is carried out by means of e.g. optical filters, i.e. externally to the filter. Also, the conversion from optical signals to one-channel (essen- tially: luminance level) signals takes place externally to the apparatus. The available signals thus represent a reduced amount of information about the three images, but the modus operandi otherwise resembles that of the sixth embodiment.
  • the eighth embodiment comprises, as shown figure 3:
  • a component, 320 for the comparison of the smoothed second (mediant colour) image signal with the signals representing the first and third (carrier colour) images and optionally modifying the mediant colour image signal m accordance with the outcome of the comparison, the modification taking the form of a further smoothing or change of amplitude or both;
  • a component, 330 for the combination of a modified first carrier colour image signal, a smoothed second (mediant colour) image signal, optionally modified, and a modified third (carrier colour) image signal, modifica- tions being controlled by the outcome of the comparisons m the component 320.
  • the following signal routes are identified by numbers: Through 340 runs the signal representing the first carrier colour image. At the branching point 391, a copy of the signal is directed towards 320. Through 341 runs the signal representing the third carrier colour and mediant colour images. At the branching point 392, a copy of the signal is directed towards 320. Through 350 runs the signal representing the second mediant colour image. Through 360 runs the signal representing the smoothed second mediant colour image. Through 370 runs the (op- tionally modified) smoothed mediant colour image signal and signals representing the outcome of the comparisons made in 320. Through 380 runs the signal representing the final, encoded image. Electronically variable stereoscopic registration can be added; but no component ef- fecting this operation is shown the figure.
  • the first embodiment of the viewing filters the range of visual light from 400nm to 700nm (light outside this range being ignored) is divided into three: the range from 400nm to 500nm (blue light) , the range from 500nm to 600nm (green light) and the range from 600nm to 700nm (red light); and a unit function, i.e. a function of constant value 1, over the full range is partitioned into three, defined as taking the value 1 on the respective sub-ranges and 0 outside.
  • the unit "block” On one of the three ranges, which will act as the mediant colour, the unit "block” is divided "horizontally", as it were; and ideal transmission rates are obtained as functions of unit value over one range, zero value over another range and a constant ("partial") value over a third range. If complementarity is not required, the partial blocks need not add up to a constant unit function; but the zeros and ones must always do ust this.
  • Figures 4, 5 and 6 show examples of idealized filter pairs obtained this way. When such ide- alized transmission functions are approximated by practical filters, the block curves will be replaced by more irregular shapes.
  • Figures 7, 8 and 9 show examples of practical filter pairs.
  • figures 4 and 7 display ([0 1/2 1], [1 1/2 0]) -pairs
  • figures 5 and 8 display ([1/2 1 0], [1/2 0 1]) -pairs
  • figures 6 and 9 display ([1 0 1/2], [0 1 l/2])-pa ⁇ rs.
  • the second embodiment of the viewing filters the division of the mediant block is "vertical".
  • Idealized transmission curves are shown figures 10, displaying a ([0 1/2 1], [1 1/2 0])-pa ⁇ r; 11, displaying a ([1/2 1 0, [1/2 0 1])- pair; and 12, displaying a ([1 0 1/2], [0 1 1/2]) -pair, respectively.
  • the first embodiment of the viewing filters allows the better colour balance of the two. Both preserve the sepa- ration property m regards to certain conventional "monochrome" stereograms (replace the "1/2" by a "0" the descriptions to obtain the corresponding non- complementary "monochrome” viewing filter descriptions) .
  • the second embodiment alongside with its use with opti- mally encoded stereograms, also allows use as barrier filters m a stereo projector:
  • stereo projectors there are two preferred embodiments of stereo projectors, as they are prescribed by the invention:
  • the filters of the tenth embodiment are mounted m a conventional stereo projector, where they replace polarizing filters.
  • Figure 13 is a schematic diagram of a conventional stereo projector, such as a slide projector or a twin digital projector system, equipped with barrier filters functionally identical to a pair of viewing filters, as these are described m the tenth preferred embodiment above.
  • Elements 1301 and 1302 represent projector units, elements 1310 and 1311 lenses or lens systems, and elements 1320 and 1321 two filters of a pair, e.g. 1320 a [0 1/2 l]-f ⁇ lter, 1321 a [1 1/2 0]-f ⁇ lter, where the "l/2"s are disjoint parts of the visible spectrum.
  • the second embodiment of a stereo projector the filters of the tenth embodiment are mounted m the light paths of a four-channel digital projector.
  • the projector is constructed like any other digital projector, be it based on LCDs, digital mirror devices (rapidly adjustable micro- mirrors) or any other technique that allows projection of full-colour images.
  • the only difference between the projector, as it is prescribed by the invention, and its conventional counterpart is that four, not three, image forming elements are required. (With projectors based on digital mirror devices (DMDs) only the time-multiplexed partitioning of the projection light need changing from a cycle of three to a cycle of four) .
  • DMDs digital mirror devices
  • the diagram figure 14 shows a light path inside part of such an apparatus, with a possible positioning of one of the filters prescribed by the invention, here thought of as a distinct from the colour separation filter employed.
  • Component 1401 illustrates a timer unit, 1410 a unit controlling intensity of light, 1420 a light source, 1430 a primary separation filter (R, G or B) , 1431 one of the filters prescribed by the invention, and 1440 a component for controlling the exit direction of the light.

Abstract

Improved methods and means for recording or encoding and displaying stereograms as single colour images for viewing through multichrome optical filters is obtained by the use of a combination of colour balanced viewing filters on the one hand, and on the other hand image encodings based on numerical optimisation techniques and colour science. Also, filters and encodings can be so selected and realized, that their application allows for a simultaneous use of some of the implements of the prior art.

Description

Title: Methods and apparatuses for encoding and displaying stereograms
The present invention relates to the encoding, or recording, and subsequent displaying and viewing of a stereogram as one colour image for viewing through a pair of filters, the filters transmitting light m the visible spectral range, each with a different transmission profile.
Background of the invention
Colour encoded stereograms are conventionally known as anaglyphs, although this term m the present context will be reserved for a sub-class of colour-encoded stereograms. This subclass of images contains the vast ma ority of images made with the prior art and will be more precisely defined shortly.
The nomenclature surrounding anaglyphs and anaglyph techniques can occasionally lead to ambiguities. Thus, the viewing filters are as a rule said to be "complementary". In colour science, two colours are called complementary if, plotted m the colour diagram, they lie on a line passing through the white point of current relevance, and on opposite sides of the white point. When greater precision is needed, complementary colours are thought of as yielding a neutral grey m additive mixing. When filters are discussed, complementarity can be tested for by shm- mg two neutral white lights through the filters separately and observing the outcome of letting the resulting lights simultaneously fall on a neutrally white surface, where they should add up to white light. When no con- straints are placed on the strengths of the two white lights, the wider definition is arrived at; and when the two white lights are required to be of equal strength, the definition is correspondingly narrowed.
Either way, it is easy to find m the prior art filter pairs which are not complementary, just as complementary filters can be found, which are of no use for anaglyph viewing, whence a better description is needed:
To allow stereopsis, i.e. the psycho-physiological response to a stereogram that results m its interpretation as a "three-dimensional image", the viewing filters must be separating. This word is here chosen to mean that there must be a first range of the visible spectrum transmitted by a first filter and excluded by a second; and a second range of the visible spectrum, not overlapping the first range and transmitted by the second filter and excluded by the first.
So-called "monochrome" anaglyphs are produced for viewing through filter pairs which are separating but not of necessity complementary, the most popular combinations being red-green or red-blue. "Monochrome" anaglyphs are m fact multichrome when considered as images m themselves. The adjective is chosen to convey the fact that no attempt is made to encode anything more than the luminance levels of the original scene, as is done m a conventional two-dimensional gray-scale image.
To avoid possible confusion surrounding the phrases "black and white (image)", "monochrome (image)" and "gray-scale (image)", the phrase "luminance level dm- age)" will be used, whenever the intended meaning is that only one channel or colour variable is represented (m the image) . The visual appearance of a luminance level image may or may not be gray. If it is not gray, only one nameable colour (i.e. one hue) is present m it, m varying intensities.
If the two ranges of a separating filter pair together cover the entire range of the visible spectrum, the fil- ters will m the present context be called "disjoint" (although the word may have other connotations). Also for convenience, the full range of the visible light can be divided into three: visible light of wavelengths less than 500 nanometers, light m the range from 500 nanome- ters to 600 nanometers, and visible light having wavelengths greater than 600 nanometers. These ranges correspond to the basic concepts of "blue light", "green light" and "red light", respectively, but it should be borne m mind that the use of these words is subjective, and also that, e.g., a "green dye" or, for that matter, a "green filter" can reflect, respectively transmit, light outside the range from 500 nanometers to 600 nanometers. The boundaries themselves can be varied some 10-20 nanometers m accordance with the task at hand. A filter pair for anaglyph viewing can be termed a (1,1) -pair, if each filter transmits light from only one of these three ranges, and a (1,2) -pair if one filter transmits light from only one range, the other from the remaining two. When it is necessary to state more precisely which ranges are transmitted and excluded, this notation can be ex¬ tended. In accordance with the common terminology, the ranges will be ordered reα-green-blue. A red-green filter pair is then denoted as a ([1 0 0], [0 1 0])-paιr, a red- cyan filter pair as a ([1 0 0], [0 1 l])-paιr etc. (As a colour, cyan is complementary to red, the word "cyan" being here used to denote an ideal mixture of green and blue. In the present parlance, a red-cyan filter pair is disjoint) . Incidentally, the numbers 1 and 0 do not literally mean "all" and "none", respectively, as no physical filter transmits all light m one spectral range and no light m another. Also, where the need arises to balance filters for complementarity m the narrower sense mentioned above, exact numbers would yield ratios appreciable different from 1:1.
A stereogram will be said to consist of two part-images, a left and a right. Terms such as "half images" are some- times seen, but it must be kept m mind that the two constituents of a (colour) stereogram are full two- dimensional (colour) images m themselves.
In the prior art, so-called "colour anaglyphs" are m- tended for viewing through disjoint filter pairs, the most popular being red-cyan filters (one exception to the use of disjoint filters is mentioned below) . In the basic form of the prior art, the images are encoded by partitioning: from the left part-image, colours and colour constituents are extracted which are transmitted by one filter m the viewing filter pair; from the right part- image, colours and colour constituents are extracted which are transmitted by the other filter; and the extracted colours and colour constituents joined to form the anaglyph. The human eye-brain responds m a numbers of ways, and in combinations of these ways, to the simultaneous presentation of distinct stimuli, one set of stimuli to each eye:
- fusion of the two perceived stimuli into one, usually recognizable as a kind of average;
- stereopsis, the combined fusion and re-mterpretation, whereby differences m the two stimuli are seen as depth information;
- rivalry, wherein fusion is impossible, yet only one actual image is seen at a time, but with the possibility of vacillations between two or more interpretations; and
diplopia, the perception of two simultaneous images
Thus, the aim of the art is to present a stereogram m such a fashion as to allow the eye-brain system to per- form fusion of colours as well as stereoscopic interpretation, and to eliminate or at least reduce rivalry where it should not appear. It is to be noted that the lustre of smooth fabrics, the gloss of polished metals and the glimmer of gems are all perceptive phenomena, produced by rivalry. These, of course, should not be suppressed. A special form of stereoscopic rivalry, commonly known as "sheen", can only be reduced, not removed: when a specific colour is transmitted by only one of the viewing filters, it may be perceived as having the right or nearly the right hue (the nameable quality, such as "pink", "lavender", "moss green"), but being materially different from other colours in the image. In reflection prints, it may e.g. seem translucent, on emissive screens it may e.g. take on a perceived luminance m excess of what the display can otherwise deliver. Apart from that, the prior art m its basic form most commonly fails to reproduce certain colours as recognizably equal to those of the original stereogram, even when complete fusion takes place; m fact, for any known embodiment of the prior art, as based on (1,2) -filter pairs and image par¬ titioning, colours can be found that do not reproduce well .
Stereopsis is usually achieved, except when the original stereogram relies on, say, texture perspective as a support of parallactic differences, and this texture perspective happens to suffer from the colour loss. Large areas of a pure colour can loose their stereoscopic ef¬ fect, if the colour is only transmitted by one of the viewing filters. Rivalry, over and above the "sheen", is typically seen where strong contrasts exist m the stereogram, or where the surrounds of the display are not sufficiently dimmed, and may m some individuals lead to discomfort (nausea, headaches etc.), sometimes ascribed to "colour bombardment". Diplopia can occur, over and above what is caused by transgression of stereoscopic parameter limits, when the recording medium, the display or the filters, or any combination of these, fail to separate the part-images of the stereogram.
The prior art
Various extensions and amendments to the art m its basic form can be found m the prior art, addressing one or more of the above problems : In Patent US 5491646, diplopia caused by insufficient separation is subjected to the countermeasure of subtracting a scaled copy of the colour values extracted from one part-image from those extracted from the other "so that said first image is entirely, or almost entirely removed from said second image", a digital counterpart of the masking technique used m printing. The prescriptions m Patent US 5491646 otherwise address the use of (1,2)- fliters .
In Patent US 4236172, the complete loss of image information m one eye m areas of pure colour is counteracted by replacing straightforward partitioning with weighted averages. The averaging concurrently serves the purpose of sharpening the perceived image, seeing that colour signals m a television are re-mterpreted from RGB- representation to a high resolution luminance signal combined with lower resolution signals carrying the chrominance information. Again, the prescriptions address the use of (1,2) -filters.
In Patent US 4217602, use is made, not of two part- images, but three. The purpose, however, is to allow non- stereoscopic, full-colour viewing, the stereoscopic ef- feet being obtained by the use of ( 1 , 1 ) -filters and thus giving a "monochrome" stereo image m the sense described above and discarding the third image. For comparison with an embodiment of the present invention, described below, it should also be noted that the tnree lenses or lens systems of Patent US 4217602 are identical.
In Patent US 4620770, use is made of two outline images ana a colour filling, rather than two definite images. The method as described pertains to hand-drawn a anaglyphs, the viewing filters are referred to simply as "3- D glasses" and appear to be intended as (1,2) -pairs, and no procedure is given for the production of specific col- ours .
In Patents FR2560398 and IT1175546, rivalry effects are eliminated by the use of disjoint filter pairs each member of which partition each of the three colour ranges into two. Extending the notation used above, such a filter pair could be referred to as a ([1/2 1/2 1/2], [1/2 1/2 1/2]), it being understood that any of the "1/2" 's is disjoint from any other. Apart from the practical difficulty of producing such filters at an acceptable cost, display of stereograms using this mode of encoding requires a conventional two-image full-colour stereo projector, or its equivalent. (Also, let it be noted m passing that the principle underlying Patent FR2560398 is already fully described m one of the standard references of stereoscopy, N.A. Valyus "Stereoscopy" , Moscow 1962, English translation published by The Focal Press, 1966) .
In Patent Application PCT/DK99/00568 , colour loss is reduced by the use of non-disjomt filter pairs. For exam- pie, a ([0 0 1], [1 1 0])-paιr (blue-yellow) is replaced by a ([0 0 1], [1 1 e])-paιr (blue-amber), where the letter "e" denotes a small, but still significant transmission rate, and the encoding by partitioning is replaced by a suitable weighted-average encoding. Then according to the descriptions m the application, loss of perceived colour, as well as sheen, can be kept at a minimum. One of the filters m the pair still transmits from only one of the tnree colour ranges, though; there is a small amount of "ghost imaging"; and at least m the blue-amber combination here given as an example, the blue filter transmits only a small portion of the actual light.
Valyus (op. cit.) p. 109, describes the use by L. Lumiere of a blue-yellow filter pair which m fact, according to the few spectral data cited, must have been a violet- yellow ([1 0 1/2], [0 1/2 l/2])-paιr. Obviously, L. Lumiere can not have attempted to display full-colour lm- ages - m fact, Valyus writes: "One of his [L. Lumiere's] images was coloured yellow, the other blue", suggesting "monochrome" anaglyphs - and there is no explicit record of further attempts with filters of this kind.
All methods m the prior art have one thing m common: they encode the images m a pomt-for-point or position- for-position fashion. The word anaglyph ("drawn twice") should be used for such colour encodings .
An argument to the effect that full-colour stereoscopy by colour encoding is fundamentally an impossibility, and the efforts of the prior art therefore m vain, might run as follows:
To produce a "monochrome" stereogram, two channels are required, one for each of the luminance level images. (A "channel" is here a single valued function, such as a luminance level function, defined on the image area. The function may take values m a circular domain, as with the hue, cp . the expression "hue circle") . To further colorate this virtual 3-D scene, a further two channels are required, typically representing the hue and saturation, respectively, of each scene point. Thus, a minimum of four colour channels is required to produce a full- colour stereogram. But a single printed, projected or displayed image contains, when no use is made of polarized light or time-multiplexing, only three channels, red, green and blue, or their re-mterpretations m terms of other three-channel colour spaces. (A colour space, often more correctly termed a "colour solid" is a representation of all possible colours assignable to an image point, for instance the available range of RGB-values. To avoid multiple meanings of the word "dimension", the phrase "three-channel colour space" is here used to denote the extension to a full image) . The failure to address this problem directly is at the heart of the various shortcomings observable m the prior art.
In opposition to the argument just given stands the observation, often made m colour science, and forming the basis of colour management, that "almost any individual colour can be made to appear as almost any other mdivid- ual colour". The stress is on "individual", because the various means of controlling the appearance of a single colour all involve controlling surrounding colours. From the argument presented above, as well as from known observations (see, e.g. Valyus op. cit. pp. 109-110) the present inventor has concluded that to achieve this m any significant measure requires use of a combination of colour balanced viewing filters on the one hand, and on the other hand image encodings that make it possible to avoid the diplopia otherwise invariably arising from the colour balancing. Also, filters and encodings should be so selected and realized, that their application allows for a simultaneous use of at least some of the implements of the prior art. Summary of the invention
The principle of the invention can be described as fol- lows: m a viewing filter pair designed as prescribed by the invention, both of the filters must transmit light m two of the ranges red, green or blue; to avoid double imaging and allow the use of optimisation and colour management, recording and encodings are no longer pomt-to- point; and display, even of existing stereograms, is colour-balanced, as far as the pertinent choice of media allows .
The present invention offers m its different aspects
- means for selecting the viewing filters to balance both the light transmission and perceived colour transmission so as to obtain a distribution of light and colours between the two eyes more even than m the prior art, while preserving a separation property;
- means for encoding existing or conventionally recorded or rendered stereograms, allowing: the use of picture elements larger than individual pixels; the use of colour management techniques as integral parts of the encoding; and the use of computational optimisation techniques, likewise as integral parts of the encoding, altogether allowing full use of a four-channel projection, when this is available, or simulating its presence on a conven- tional three-channel display;
- means for directly recording a stereogram m the form of one co our-oalanced colour imaqe and witnout the for- ation of artefacts that would otherwise be visible m stereoscopic viewing as double images;
- means for displaying optimally encoded as well as con- ventionally recorded stereograms m the form of one colour-balanced colour image and without the formation of artefacts that would otherwise be visible m stereoscopic viewing as double images.
With reference to the division of the visible spectrum into three ranges, red, green and blue, let two of these be designated "the carrier colour ranges" and the third "the mediant colour range". A filter pair as prescribed by the invention must then be a "(3/2, 3/2) -pair" m the sense that the filters must be separating with respect to the carrier colour ranges; but both filters must transmit the mediant colour range. This latter can take the form of a joint transmission m the entire mediant colour range; or each filter can transmit a part of this range. Transmission m the mediant colour range may involve some reduction, relative to transmission m the pertinent carrier colour range. Henceforth, where no misunderstanding can arise, the phrase "colour range" will be replaced by the simpler "colour". In a stereogram, encoded as pre- scribed by the invention, three separate images can then be recognized: one m each of the carrier colours and one in the mediant colour.
When the viewing filters are so chosen that, apart from their respective transmittance and mutual exclusion of carrier colour range, they each transmit one part of the mediant colour range and exclude the part transmitted by the other, the same filters can be used as barrier fil- ters m a display device. If this device is a conventional stereoscopic projector, the filters can thus take the place of polarizing barrier filters. In a digital projector, the filters can again take the place of polar- lzing filters, but there is the further advantage that m the left part-image, no constituent corresponding to the carrier colour of the right part-image need ever be formed (or it can be suppressed); and vice versa. Such a digital projector therefore, m a sense, projects four colour channels: the carrier colour associated with the left part-image and transmitted by the left-image filter; the mediant colour associated with the left part-image, but projected only m the sub-range transmitted by the left-image filter; the mediant colour associated with the right part-image, but projected only m the sub-range transmitted by the right-image filter; and the carrier colour associated with the right part-image and transmitted by the right-image filter. In this case, the requirement of four perceptual colour channels is met directly and can be fully utilized by an optimising encoding of the stereogram, while m the remaining encodings described below, primarily aimed at conventional three- channel displays, the impression of a four-channel display is created by colour management.
Fully developed, the encoding of a stereogram as prescribed by the invention involves an optimisation proc¬ ess. Generally, the precise effects of an implicitly de¬ fined process cannot be predicted (since otherwise the prediction would be an explicit process replacing the im¬ plicit one) , but some qualitative features of this opti¬ misation are evident: The resulting mediant colour image, considered as an image m itself, is visible to both eyes and can therefore not be a single, sharp image, as this would cause diplopia. It must therefore, considered as an image m itself, appear blurred or double (or both) , these effects being strongest in areas which represent scene objects far from the zero parallax depth.
It follows that the two carrier colour images, considered as images m themselves, must, at least to some extent, compensate for any loss of sharpness caused by this blurring of the mediant colour image. Thus, the amplitudes of the left carrier colour image will be enhanced, wherever the amplitude of the mediant colour image is smaller than that of the mediant colour component of the original left part-image; and it will be dampened, wherever the amplitude of the mediant colour image is larger than that of the mediant colour component of the original right part- image. The same thing holds, mutatis mutandis, for the right carrier colour image.
Also, as an integral part of the encoding, double images will be removed or reduced, seeing that contributions to the luminance variation from the original right part- image to the encoded left part-image will become part of the overall luminance variation of the encoded left part- image. Again, the same holds, mutatis mutandis, for the right part-image.
Finally, among the possible changes that will produce the desired effects on the luminance levels, an optimisation process will, when it is operating on a so-called "cost function" that also takes into account the final per- ceived colours, select those that cause the smallest changes (and hence, if possible: no changes) m the fused colours, compared with the fused colours perceived m the original stereogram, when this is stereoscopically viewed.
The re-distribution of luminance between the three images thus serves to obtain a balance between the four distinct requirements :
- that the luminance impression of the part-image received by the left eye when observing the encoded stereogram through the assigned left filter resembles (a scaled version of) the luminance impression received when ob- serving the original left part-image;
- that the luminance impression of the part-image received by the right eye when observing the encoded stereogram through the assigned right filter resembles (a scaled version of) the luminance impression received when observing the original right part-image;
- that the fused hues resemble those observed m the original stereogram as stereoscopically viewed; and
- that the fused saturations resemble those observed m the original stereogram as stereoscopically viewed.
To obtain this, the encoding must replace some amount of one or both carrier colour with mediant colour, and vice versa, and this replacement can not be performed on a pomt-by-point basis, as this would create visible dis- continuities that would give rise to both diplopia and abrupt colour casts.
Arguably, the simplest instantiation of the encoding, here expressed in photographical terms, is obtained by first photographically recording, along three parallel lines of sight, three (luminance level) images, the middle one having a restricted depth of field m the sense of being visually sharp only m the distance range m- tended for zero parallax m the final stereogram, then encoding the outermost two m the respective carrier colours and the middle one in the mediant colour, and finally mounting the three m register, i.e. with the zero parallax motif parts coalescent and homologous points horizontally aligned, to form a full-colour image. As an alternative to recording gray scale images, colour separation filtered images can be recorded. Note, that for direct recording, the lens system must allow m-place registration through a shift of the image frame. A re- cording process of this kind can be accomplished by an optical apparatus, but it should be noted that such an apparatus performs no amplitude enhancements of the carrier colour constituents. Note, also, that most analog films will give rise to double (actually: triple) images owing to insufficient spectral separation. The process can also be accomplished by an electronic filter, forming part of e.g. an electronic camera or image recorder. Digital image recorders can avoid the double imaging caused by spectral separation failures; and an electronic filter can perform enhancements of the carrier colour constituents and indeed other cross-constituent operations, as they are described m the following. The description of the image encoding as prescribed by the invention can proceed m steps, each resulting instantiation being a variation over or an improvement of this theme:
- Leaving the world of analog photography: the middle image need not be directly recorded, but can be computed as an average of the mediant colour "planes" of the two outer images, each of which must then be recorded in at least two colours, the mediant colour and the relevant carrier colour. The averaging can be simple (half the sum), weighted using pre-determmed weights, or adap- tively weighted, taking into account the fact that areas which are low in one of the carrier colours provide less room for subsequent colour manipulations.
If the averaging itself is a simple point-to-point weighted summation, the optical blurring can be simulated by a smoothing process. The smoothing operator can be a simple "sliding strip" technique, one or two-dimensional, or a genuine approximation to the point spread function of a lens with finite (as opposed to point-shaped) aperture. The local strength of the smoothing can vary according to any measure of local parallactic separation. Assuming that the stereo registration has been decided, this can be put m simpler terms: if there is very little difference between the mediant colour contents (or, alternatively, the gray scale contents) of the left and right images m and around a given position, the mediant colour should be only gently smoothed, in and around that position. - Once an averaged and smoothed (henceforth just referred to as: "averaged") mediant colour image is available, the relationships between each carrier colour and the mediant colour can be reconsidered: to reduce loss of sharpness, which may occur when, m and around a given location, the averaged mediant colour differs from the mediant colour image of the one or the other of the part-images, the relevant carrier colour may be proportionally enhanced or reduced, m and around that location.
- Such a replacement may also be used to remedy double image phenomena: if, owing to the spectral imperfections of recording media, display media or filters, an "echo" of the left part-image is visible to the right eye or vice versa, the one carrier colour image may be modified by subtraction of a scaled version of the other carrier colour image. (Lest it may seem that this procedure leads to an infinite regress, it should be noted that the problem can be stated as two simultaneous equations m two unknowns; that the repeated procedure, if m fact performed, has the character of a "relaxation solution" of this problem; and that the smallness of the scaling factors implies a rapid convergence) .
- The actual measures of equilummant replacement will not be simply linear (proportional) as expressed m terms of colour coefficients (typically: "RGB"-values m digital representation) , not even after taking into account the "gamma" of the medium at hand. (The "gamma" is the coefficient m the logarithmic/exponential relationship between e.g. image coefficients and CRT voltages, more generally: between numerical representation and light effect; . They are therefore better cased on spectrally measured colour gamuts. In practice, conversion back and forth is typically performed by table look-ups and interpolation. This is no different from all other colour computations, and can be built into the stereogram encoding.
- Rather than basing the smoothing strength on estimates of parallactic separation of the left and right part- images m and around a given location, it can be based directly on a correspondence map of the stereogram. This latter is a representation, typically m the form of a gray scale image, of the separation at any location, of the point at that location m one of the part-images and its homologous point m the other part-image. (Homologous points are images of one and the same scene point. Their separation is geometrically related to the depth of the scene point) . Algorithms for solving the correspondence problem exist m great number m the arts of photogram- metry and machine vision.
- With the help of an algorithm for solving the correspondence problem, the stereogram can be artificially de- saturated with little loss of image quality: if the saturation of colours of distant areas is reduced, they will be more easily encoded without causing artefacts of dis- continuities m, or mis-coloration of, areas closer to the foreground. As with the adaptive averaging, this is again a question of providing room for colour re¬ distribution .
Finally, most or all of the above can be combined. But to make the description of a combined algorithm understand¬ able, one technical detail must be inserted: In a conventional stereo viewer (e.g. a Brewster-viewer) , the two part-images of a given stereogram are placed side by side and optically brought to seem coalescent. In an anaglyph, or a stereogram encoded by the present lnven- tion, two images actually coalesce by construction. Hence, any given image area serves two purposes simultaneously: it is the representation of an area m the original left part-image - and a representation of a different area m the original right part-image. Hence, to a given area m a colour-encoded stereogram, considered as seen by the left eye, a homologous area can be found. For convenience, the stereogram is assumed to be an endo- stereogram (all scene points are perceived as lying behind the image plane), so the homologous area lies to the right of the given area. When next the latter area is considered as seen by the left eye, a new homologous area can be found further to the right, etc. A model for what is actually perceived m a colour encoded stereogram must represent such chains of homologous areas, typically m the form of lists of selected representative homologous point pairs and their surroundings. Colour fusion will be between the colours seen m an area and its homologous counterpart. The areas must be found m (and referred back to) the original stereogram.
Now, given: a norm or metric on a device-independent col¬ our space; an algorithm to solve the correspondence problem; a model for colour fusion; spectral measurements and models to convert device and image colour values into de- vice-independent values; an algorithm for numerical opti¬ misation; a viewing filter pair chosen as prescribed by the invention; and a stereogram - the image encoding pre¬ scribed by the invention: solves the correspondence problem for the original stereogram, representing the solution as a list or lists of homologous areas; - computes the perceived fused colours m these areas, according to the fusion model; sets up the optimisation problem of encoding the stereogram m such a fashion as to express that what is actually perceived (as fused colours m the same areas) m the encoded stereogram, when this is observed through the viewing filters, has the smallest deviation m terms of the norm or metric on the colour space from the fused colours m the original stereogram;
- solves the optimisation problem using the numerical method, arriving at the encoding as the solution of this problem; and
- computes the image or device representation from the resulting device-independent colour representation.
It remains to be mentioned, that display of the recorded or encoded images can be made using any conventional (colour) display medium, including the media of prints or colour photos. If circumstances allow, though, use can advantageously be made of special stereoscopic projectors (designed as prescribed by the invention) accommodating coloured filters, which must then be functionally matched by the viewing filters.
In the following a short vocabulary will be given to more precisely set out the terms and their intended meaning used m the context of the present application, m the description as well as m the claims. A part of the spectrum will be called a "range" if it forms a single, connected interval. A range, including the complete visible range, can be "partitioned", meaning divided into smaller ranges, which together exhaust the original. A typical partitioning of the entire visible range has been mentioned previously: blue light with wavelengths less than to 500nm, green light with wavelengths from 500nm to 600nm and red light with wavelengths greater than 600nm.
It is customary to think of a colour image as consisting of "components", such as its R-, G- and B-components, and this is done m the following, even if the actual display medium uses "primaries" of a different composition than suggested by the partitioning considered. An image will be said to have components "after" each of a given set of ranges, meaning that it could be composed from these components if ideal primaries existed, each emitting light at unit strength over a range and at zero strength out- side it. Conversely, adding up the spectral distributions of such components is one way of "combining" the components into a resulting image.
These ideal operations have their well-known physical counterpart: separating an RGB-signal into an R-, a G- and a B-signal (whether analog or digital) is conceptually equivalent to passing light through "colour separation filters", such as the photographic filters mentioned above. In the case of physical light, practical sense must be made of some of these idealizations: an ideal clear filter would have a transmission rate of 1 everywhere; a physical filter can be thought of as having a "transmittance" , a function taking values less than or equal to 1 everywhere on the range of visible light or a sub-range of it. Rather than referring to point values, the transmittance over a range shall mean the area under the transmission rate function graph divided by the width the range; and a filter will be said to "transmit a range", if its transmittance over that range is larger than 0.06 (6C), and to "exclude a range" if its transmittance over that range is less than 0.06. (This value corresponds to a damping of four photographic "stops" and marks a practical limit at which e.g. the presence of "ghost images" begins to cause diplopia) . Two filters will be said to have a "significant common transmission" of a range, if the common area of their transmission rate functions restricted to that range is no less than 10 of either area.
From the realm of stereoscopy: two points m two part- images of a stereogram are said to be "homologous" if they are the projected images of one and the same point m the scene depicted. The extension to areas is obvious. "Fusional areas" are usually considered to be such areas, the images of which jointly fall on the foveas of an observer's eyes; but the definition can be slackened or tightened as the application at hand requires, the mean- mg being always that the areas are homologous, and that their contents are candidates for stereoscopic fusion. A "correspondence map" is defined elsewhere m the text and it remains only to note, that it is understood that homologous point pairs are at least approximately brought to lie on parallel lines ("horizontal lines"), and that such a map may take on negative values (negative parallaxes , if part of the scene depicted lies front of the image plane used in the depiction. If the left and right part-images of an alleged stereogram are actually identical, the scene depicted is a flat object, parallel to the image plane, and its correspondence map is constant, possibly everywhere zero.
Finally: an image can be considered a two-dimensional signal, and concepts such as discrete Fourier transforms applied to it. Using such a transform one may define frequency ranges and their energy contents. Reference is made to the literature on the subject. Only the idea of "smoothing" an image by removing or reducing some of its "energy" m the "high frequency range" is needed m the following. Such a description is valid, even if the actual smoothing is carried out by other means.
Description of the drawings
In the following different preferred embodiment of the present invention will be described in more detail with reference to the figures in which:
Figure 1 is a diagram showing the signal routes and operational units of an apparatus that performs the encoding of a stereogram as prescribed by the invention;
figure 2 is a diagram showing the functional elements an optical apparatus that performs the encoding of a stereogram as prescribed by the invention;
figure 3 is a diagram showing the signal routes and operational units of an apparatus that performs the encoding of a stereogram as prescribed by the invention; figures 4 to 6 show idealized transmission rates of viewing filter pairs as prescribed by the invention;
figures 7 to 9 show transmission rates of physical view- mg filter pairs designed as prescribed by the invention;
figures 10 to 12 show idealized transmission rates of filter pairs for use, as viewing filters, and, m the projection apparatuses prescribed by the invention, si- multaneously as barrier filters;
figure 13 is a diagram showing the application of a filter pair with spectral properties derived from the ideal transmission rates of one of figures 10 to 12 as barrier filters a conventional stereo projector; and
figure 14 shows one light ray path and a possible positioning of one of the filters of figure 10 to 12, jointly with a colour separation filter, m a projector as pre- scribed by the invention.
Detailed description of preferred embodiments of the invention
In a first preferred embodiment of the image encoding method according to the invention, a stereogram is given as two part-images digital form, represented by their "RGB" value arrays (as is customary m e.g. computer graphics). To make the following description specific, the left carrier colour is exemplified by B (blue), the right carrier colour by G (green) and the mediant colour by R (red), other choices being immediately obtainable by exchange of letters. The R components of the left and right images are then halved, added up and subjected to a smoothing m the form of a "sliding strip" averaging with a weight function m the form of a "Gaussian bell" (both terms defined and described detail m standard refer- ences on numerical analysis and image analysis) . The encoded image is obtained as the combination of this averaged R component with the B component from the left image and the G component of the right image. The simplicity of this embodiment is evident, but it is only truly applica- ble to stereograms with low contrasts and restricted depth .
The use of definite colour letters is maintained everywhere the following, with the same proviso that other combinations can be obtained by exchange of letters. Also, the "RGB" readings must be considered as transferred to additive values using the pertinent "gamma" relation, prior to all computations, and back again prior to display of the encoded image.
In a second preferred embodiment of the image encoding, an approximate correspondence map is first obtained, for example by simply computing the point-wise differences m luminance levels everywhere over the two part-images of the stereogram (this is crude, but remarkably effective for its present use) . An adaptive averaging of the two R- components of the part-images is then computed: First, a point-wise sum is formed of the minimum of the two R- values with a weighted sum of the two differences between the minimum and the actual values (thus, m any given point, one or the other of the contributions is zero) . The weights are so chosen, relative to the luminances of reα, green and blue the αisplay medium and to a selec- tion of test images, so that colour casts m the test images are found acceptable by visual inspection. The choice of weights is thus empirical (psycho- physiological ) . The result is smoothed, as described with reference to the first embodiment, with the smoothing strength now coupled to the values of the correspondence map. The coupling is again determined by visual inspection, m order to avoid diplopia. The encoded image is obtained as the combination of this averaged R component with the B component from the left image and the G component of the right image.
In a third preferred embodiment of the image encoding, encoding proceeds as m the second embodiment, but the B and G components are further corrected to eliminate or nearly eliminate luminance variations ("ghost images") caused by the failure of the recording medium or of the viewing filters relative to the display medium or both to preserve ideal spectral separation.
In a fourth preferred embodiment of the image encoding, a genuine, if not exact, correspondence map is first obtained (by any method of choice) and the saturation of colours m the part-images of the stereogram reduced, m- creas gly with increasing depth, colours of the nearest points remaining unreduced. Encoding proceeds as m the third embodiment, except that the degree of smoothing is now more precisely controllable, a more precise corre¬ spondence map being available.
It should be noted here, that for artificially generated (digitally rendered) stereo images, an exact correspondence map can be generated m the course of the rendering process, either directly by the expedient of monitoring the projection of scene points, or indirectly from the so-called depth buffers of the part-images.
A fifth preferred embodiment of the image encoding requires a veritable battery of auxiliary tools:
- a method for determining a correspondence map of a stereogram;
- a choice of a size of fusional area(s) and with this, and with the correspondence map, a partitioning of the stereogram into chains of homologous areas;
- a device-independent colour space with a metric (colour distance) defined on it (several such are described colour science) ; - an extension of the metric, such as an integration or weighted summation of point-wise values of the metric, to an error measure allowing comparison of colours of fused homologous areas, as these are seen m the original stereogram, and m the encoding, respectively; - a model for colour fusion (with additive mixture as the very simplest); and
- a numerical optimisation method (many examples of which can be found m the art of numerical analysis) .
The encoding, as prescribed by the invention, is then found as the solution to the optimisation problem of reducing the error, as determined by the error measure, of the fused colours seen m the encoded image, compared with the fused colours m the stereogram, as these are determined according to the colour fusion model, acting over homologous areas m the chains partitioning the stereogram. There are three preferred embodiments of the invention m the form of apparatuses aiding the recording of stereograms :
A sixth preferred embodiment of the invention, the first preferred embodiment of the apparatus aiding the recording of stereograms, takes the form of an electronic filter converting three incoming full-colour (or: a two- colour, a one-colour and a two-colour) image signals to one outgoing full-colour image signal. This embodiment comprises, as shown m figure 1:
- a component, 101, for separating the partial signal representing the first carrier colour and the mediant colour from the first image signal; - a component, 102, for separating the partial signal representing the mediant colour from the second image signal;
- a component, 103, for separating the partial signal representing the second carrier colour and the mediant colour from the third image signal;
- a component, 110, for the pre-smoothmg (low pass filtering) of the mediant colour image signal;
- a component, 120, for the comparison of the smoothed second mediant colour image signal with the signals rep- resenting the first and third mediant colour images and with the signals representing the first and third carrier colour images (and optionally modifying the second mediant colour image signal m accordance with the outcome of the comparison, the modification taking the form of a further smoothing or change of amplitude or both) ; and
- a component, 130, for the combination of a modified first carrier colour image signal, a smoothed second mediant colour image signal (optionally modified) and a modified third carrier colour image signal, modifications being controlled by the outcome of the comparisons m the component 120.
It should be noted that if the original incoming signals are made, by some means external to the apparatus, to represent a conforming two-colour, a one-colour, and a two-colour image, respectively, then components 101, 102 and 103 are passive and, m a sense, superfluous. In the figure, the following signal routes are identified by numbers: Through 140 runs the signal representing the first carrier colour and mediant colour images. At the branching point 191, copies of both signals are directed towards 120. Optionally, only the carrier colour signal proceeds from there to 130. Through 141 runs the signal representing the third carrier colour and mediant colour images. At the branching point 192, copies of both signals are directed towards 120. Optionally, only the car¬ rier colour signal proceeds from there to 130. Through 150 runs the signal representing the second mediant colour image. Through 160 runs the signal representing the smoothed second mediant colour image. Through 170 runs the (optionally modified) smoothed mediant colour image signal and signals representing the outcome of the com- paπsons made m 120. Through 180 runs the signal repre¬ senting the final, encoded image.
It should be noted that it is not difficult m the sixth preferred embodiment described above to recognize the elements of the second and third embodiments. What was said m their descriptions about choices of weights and elimination of "ghost" images holds true m their elec¬ tronic counterpart. A seventh preferred embodiment of the invention, the second preferred embodiment of the apparatus aiding the recording of stereograms, takes the form of an optical adaptor converting three visual image signals to one outgoing visual image signal. The embodiment comprises, as shown m figure 2 :
- optical colour separation filters 201, 202 and 203. These can for instance be the filters most frequently used other photographic colour separation work, Kodak Wratten 25 (red) , Kodak Wratten 58 (green) and Kodak Wratten 47B (blue) m some order. (The spectral properties of these filters are described m Kodak's technical references, and also elsewhere) ; - lenses or lens systems, 210, 211 and 212, with 211 having a reduced depth of field relative to 210 and 212 (which are identical);
- mirrors 220 and 221; and
- prismatic mirrors 230, allowing passage of an lmage- forming light ray (bundle) from 211 and reflecting an image-forming light ray (bundle) from each of 220 and 221 such a fashion as to produce a combined image.
As described so far, the device would be of fixed stereo- scopic registration. A coupling (not shown separately in the drawing) between the lens system of the camera, on which the adaptor is mounted, and e.g. the lenses or lens systems 210 and 212 can be added to allow variable stereoscopic registration.
An eighth preferred embodiment of the invention, the third preferred embodiment of the apparatus aiding the recording of a stereogram, takes the form of an elec- tronic filter to be mounted m a three-lens camera, m which colour separation is carried out by means of e.g. optical filters, i.e. externally to the filter. Also, the conversion from optical signals to one-channel (essen- tially: luminance level) signals takes place externally to the apparatus. The available signals thus represent a reduced amount of information about the three images, but the modus operandi otherwise resembles that of the sixth embodiment. The eighth embodiment comprises, as shown figure 3:
- a component, 310, for (pre-) smoothing the second (mediant colour) image signal ;
- a component, 320, for the comparison of the smoothed second (mediant colour) image signal with the signals representing the first and third (carrier colour) images and optionally modifying the mediant colour image signal m accordance with the outcome of the comparison, the modification taking the form of a further smoothing or change of amplitude or both; and
- a component, 330, for the combination of a modified first carrier colour image signal, a smoothed second (mediant colour) image signal, optionally modified, and a modified third (carrier colour) image signal, modifica- tions being controlled by the outcome of the comparisons m the component 320.
In figure 3 the following signal routes are identified by numbers: Through 340 runs the signal representing the first carrier colour image. At the branching point 391, a copy of the signal is directed towards 320. Through 341 runs the signal representing the third carrier colour and mediant colour images. At the branching point 392, a copy of the signal is directed towards 320. Through 350 runs the signal representing the second mediant colour image. Through 360 runs the signal representing the smoothed second mediant colour image. Through 370 runs the (op- tionally modified) smoothed mediant colour image signal and signals representing the outcome of the comparisons made in 320. Through 380 runs the signal representing the final, encoded image. Electronically variable stereoscopic registration can be added; but no component ef- fecting this operation is shown the figure.
There are two preferred embodiments of the viewing filters, as they are prescribed by the invention:
In a ninth preferred embodiment of the invention, the first embodiment of the viewing filters, the range of visual light from 400nm to 700nm (light outside this range being ignored) is divided into three: the range from 400nm to 500nm (blue light) , the range from 500nm to 600nm (green light) and the range from 600nm to 700nm (red light); and a unit function, i.e. a function of constant value 1, over the full range is partitioned into three, defined as taking the value 1 on the respective sub-ranges and 0 outside. On one of the three ranges, which will act as the mediant colour, the unit "block" is divided "horizontally", as it were; and ideal transmission rates are obtained as functions of unit value over one range, zero value over another range and a constant ("partial") value over a third range. If complementarity is not required, the partial blocks need not add up to a constant unit function; but the zeros and ones must always do ust this. Figures 4, 5 and 6 show examples of idealized filter pairs obtained this way. When such ide- alized transmission functions are approximated by practical filters, the block curves will be replaced by more irregular shapes. Figures 7, 8 and 9 show examples of practical filter pairs. In the notation introduced m the outline of the background of the invention, figures 4 and 7 display ([0 1/2 1], [1 1/2 0]) -pairs, figures 5 and 8 display ([1/2 1 0], [1/2 0 1]) -pairs and figures 6 and 9 display ([1 0 1/2], [0 1 l/2])-paιrs.
In a tenth preferred embodiment of the invention, the second embodiment of the viewing filters, the division of the mediant block is "vertical". Idealized transmission curves are shown figures 10, displaying a ([0 1/2 1], [1 1/2 0])-paιr; 11, displaying a ([1/2 1 0, [1/2 0 1])- pair; and 12, displaying a ([1 0 1/2], [0 1 1/2]) -pair, respectively.
The first embodiment of the viewing filters allows the better colour balance of the two. Both preserve the sepa- ration property m regards to certain conventional "monochrome" stereograms (replace the "1/2" by a "0" the descriptions to obtain the corresponding non- complementary "monochrome" viewing filter descriptions) . The second embodiment, alongside with its use with opti- mally encoded stereograms, also allows use as barrier filters m a stereo projector:
There are two preferred embodiments of stereo projectors, as they are prescribed by the invention:
In an eleventh preferred embodiment of the invention, the first embodiment of a stereo projector, the filters of the tenth embodiment are mounted m a conventional stereo projector, where they replace polarizing filters. Figure 13 is a schematic diagram of a conventional stereo projector, such as a slide projector or a twin digital projector system, equipped with barrier filters functionally identical to a pair of viewing filters, as these are described m the tenth preferred embodiment above. Elements 1301 and 1302 represent projector units, elements 1310 and 1311 lenses or lens systems, and elements 1320 and 1321 two filters of a pair, e.g. 1320 a [0 1/2 l]-fιlter, 1321 a [1 1/2 0]-fιlter, where the "l/2"s are disjoint parts of the visible spectrum.
In an twelfth preferred embodiment of the invention, the second embodiment of a stereo projector, the filters of the tenth embodiment are mounted m the light paths of a four-channel digital projector. The projector is constructed like any other digital projector, be it based on LCDs, digital mirror devices (rapidly adjustable micro- mirrors) or any other technique that allows projection of full-colour images. The only difference between the projector, as it is prescribed by the invention, and its conventional counterpart is that four, not three, image forming elements are required. (With projectors based on digital mirror devices (DMDs) only the time-multiplexed partitioning of the projection light need changing from a cycle of three to a cycle of four) . With the previously introduced proviso of letter exchangeability, if a conventional projector uses three units, one for each of R, G and B, or a DMD runs a cycle of R,G,B, these must be expanded to RGGB and RGBG, respectively. Then the first pair, say, of these is used to form the left image, the light being further passed through a [0 1/2 1] -filter; and the second pair is used to form the right image, the light being passed through a [1 1/2 0] -filter. If the filters are complementary, an ordinary image will be projected as it is; and a stereo image projected this way can be viewed with an identically, or functionally equivalent viewing filter pair. The special optimisation encoding technique of the fifth preferred embodiment applies directly to this situation, since only the fusion model need differ from the three-channel case. But the option of actually encoding the image m four channels is evident, only it has no application outside the use of this class of projectors. The diagram figure 14 shows a light path inside part of such an apparatus, with a possible positioning of one of the filters prescribed by the invention, here thought of as a distinct from the colour separation filter employed. Component 1401 illustrates a timer unit, 1410 a unit controlling intensity of light, 1420 a light source, 1430 a primary separation filter (R, G or B) , 1431 one of the filters prescribed by the invention, and 1440 a component for controlling the exit direction of the light.
It will be appreciated that many modifications of the invention are possible without departing from the scope of the invention as defined m the accompanying claims.

Claims

1. A method for encoding a stereogram as a single colour image A, the stereogram comprising a first part- image S and a second part-image D, the method comprising the steps of:
- partitioning of the visible part of the spectrum into a first range K, a second range L and a third range M: - separating from the part-image D a component image DK after K and a component image DM after M;
- separating from the part-image S a component image SL after L and a component image SM after M;
- forming a modified image AM after M from SM and DM, for instance a smoothed average of SM and DM;
- forming a modified image AK after K from SK, for instance a sum of SK and a luminance correction CK;
- forming a modified image AL after L from DL, for instance a sum of DL and a luminance correction CL; and
- combining AK, AL and AM to obtain the single colour image A.
2. A method for encoding a stereogram as a single colour image A as defined claim 1, the method comprising the further steps of:
- separating from the part-image D a component image DK after K and a component image DM after M;
- separating from the second part-image S a com- ponent image SL after L and a component image SM after M;
- forming an average image TM after M from SM and DM; - applying a smoothing operator OS to TM to obtain a smoothed image AM after M from TM;
- forming AK by letting AK equal SK;
- forming AL by letting AL equal DL; and - combining AK, AL and AM to obtain A.
3. A method for encoding a stereogram as a single colour image A as defined m claim 2, the method comprising the further steps of: - forming from S and D an approximate correspon¬ dence map C of the stereogram and associating points of C with points of TM;
- allowing the smoothing operator OS to have a controllable, variable effect depending on a single pa- rameter p, such that the effect of OS is smallest, when p is equal to zero, and such that the effect of OS grows monotonically with the numerical value of p; and
- controlling the effect of OS on a region RM of TM by taking as a value PR of p, a numerical value of C m RM.
4. A method for encoding a stereogram as a single colour image AAA as defined m claim 3, the method com¬ prising the further steps of: - adding to A a component image ADK after K m the form of scaled copy, for instance a negatively scaled copy, of the values of DL, to obtain an image AA; and
- adding to AA a component image ASL after L m the form of scaled copy, for instance a negatively scaled copy, of the values of SK, to obtain AAA.
5. A method for encodmg a stereogram as a smgle colour image AAA as defined claim 4, the method comprising the further steps of:
- modifying, after the formation of the corre- spondence map C, and prior to the remaining steps of claim 4, and as part of the steps of separating SL and SM from S and DK and DM from D, the part-images S and D by applying a saturation-diminishing operator OSD to S and D respectively, allowing the operator OSD to have a con- trollable, variable effect depending on a single parameter p, such that the effect of OSD is smallest, when p takes a value equal to the minimal value found C, and such that the effect of OS grows monotonically with the value of p, controlling the effect of OSD on S and D re- spectively by taking as a value P of p, m a region RS of
5, respectively a region RD of D, a value of C formed from RS and RD.
6. A method for encoding a stereogram as a single colour image AAA as defined m any of claims 1-5, wherein the component images AM, AK and AL of the method as described m claim 1 being implicitly defined as solution elements of an optimisation problem P, formulation and solution of P comprising the steps of: - determining a correspondence map C from S and
D;
- choosing a size Z of fusional area(s) and with this and C, partitioning the stereogram into a collection H of chains of homologous areas; - choosing a αevice-independent colour space with a colour distance defined on it, for instance the L*u*vψ space; - defining an extension V of the colour distance, for instance as the sum of colour distances of regularly spaced points within a fused area, to allow comparison of fused colours m homologous areas of the original stereogram with fused colours m the same homologous areas m an encoding; defining a further extension of the error measure, such as summation over the fused areas m a chain, to allow computation of a total error over a chain of fused homologous areas;
- choosing a model for colour fusion, for instance additive mixture;
- defining P as the problem of modifying AM, AK and AL to reduce to a minimum the error, as determined by the error measure, of the fused colours seen m A, compared with the fused colours in the stereogram, as these are determined according to the colour fusion model, acting over homologous areas m the collection H of chains partitioning the stereogram; and - solving P for AM, AK and AL .
7. An apparatus for aiding m encoding of a stereogram, the apparatus comprising:
- means for separately receiving a representa- tion of a left part-image S, a representation of a central part-image C and a representation of a right part- image D;
- means for effecting a separation from S of a component SK; - means for effecting a separation from C of a component CM;
- means for effecting a separation from D of a component DL; - means for effecting a modification of CM, for instance a smoothing operation, to form a component AM;
- means for effecting a modification of SK, for instance an amplitude correction, to obtain a component AK;
- means for effecting a modification of DL, for instance amplitude correction, to obtain a component AL; and
- means for collecting AK, AM and AL into the representation of one image.
8. An apparatus as defined m claim 7, the apparatus further comprising:
- means for accommodating a first colour separa- tion filter FFK transmitting a first spectral distribu¬
- means for accommodating a second colour separation filter FFL transmitting a second spectral distribution LL; - means for accommodating a third colour separation filter FFM, transmitting a third spectral distribu¬
- means LSK, such as a first lens or lens sys¬ tem, for image formation from light that has passed through FFK, LSK having a first depth-of-fleld function F;
- means LSL, such as a second lens or lens sys¬ tem, for image formation from light that has passed through FFL, LSL having a second depth-of-fleld function G being significantly different from F;
- means LSM, such as a third lens or lens system, for image formation from light that has passed through FFM, LSM having a third depth-of-fleld function H equalling F; and
- means, for instance m the form of a combination of mirrors and prisms, for gathering images formed by LSK, LSL and LSM into one image.
9. An apparatus as defined m claim 7, the apparatus providing upon reception of S, C and D as defined m claim 7, an electronic representation EA of a single lm- age A, the apparatus comprising:
- means for receiving an electronic representa
Figure imgf000043_0001
- means for receiving an electronic representa
Figure imgf000043_0002
- means for receiving an electronic representa
Figure imgf000043_0003
- means for partitioning from ES a first contribution ESK and a second contribution ESM;
- means for partitioning from ED a first contri- bution EDL and a second contribution EDM;
- means for partitioning from EC a contribution ECM - means for forming a first contribution EAK to EA, EAK being essentially equal to ESK;
- means for forming a second contribution EAL to EA, EAL being essentially equal to EDL;
- means for forming a third contribution EAM to EA after MM, EAM being essentially an average of ESM, EDM and ECM, the average being optionally weighted and op- tionally subjected to a smoothing; and
- means for assembling EA from the contributions EAK, EAL and EAM.
10. An apparatus as defined in claim 7 for transmitting an electronic representation EA of a single image A composed from a first part-image S and a second part- image D, the apparatus further comprising: - means for receiving an electronic representation ESK of a component of S;
- means for receiving an electronic representation EDL of a component of D;
- means for receiving an electronic representa- tion ECM of a component of C;
- means for forming a first contribution EAK to EA, EAK being essentially equal to ESK;
- means for forming a second contribution EAL to EA, EAL being essentially equal to EDL; - means for forming a third contribution EAM to
EA after MM, EAM being essentially a smoothing of ECM; and
- means for assembling EA from the contributions EAK, EAL and EAM.
11. An optical filter pair for displaying or viewmg a stereogram, the filter pair comprising a first filter FS and a second filter FD, such that relative to a partitioning of the visible part of the spectrum into a first range K, a second range L and a third range M, m which each of the parts K, L and M corresponds essentially to one of the named colours "red", "green" and "blue":
- FS transmits K and transmits M and excludes L;
- FD transmits L and transmits M and excludes K.
12. An optical filter pair for displaying or viewing a stereogram as defined m claim 11, wherein the filter pair further nas a significant common transmission of M.
13. An optical filter pair for displaying or viewing a stereogram as defined in claim 11, further comprising a partitioning of MM into a range MS and a range MD, such that:
- FS transmits MS and excludes MD; and
- FD transmits MD and excludes MS.
14. An optical filter pair for displaying or viewing a stereogram as defined in claim 11, wherein the filter
FS has a transmittance TSK of K, a transmittance TSL of L and a transmittance TSM of M, the filter FD having a transmittance TDK of K, a transmittance TDL of L and a transmittance TDM of MM, such that: - TSM is no greater than 150% of TSK and no less than 10% of TSK;
- TDM is no greater than 150% of TDL and no less than 10% of TDL;
- TSM is no greater than 1000% of TDM and no less than 10 of TDM;
- TSL is no greater than 6. of TSK; and
- TDK is no greater than 6-0 of TDL.
15. An optical filter pair for displaying or viewing a stereogram as defined in claim 14, the filter pair further having a significant common transmission of M.
16. An optical filter pair for displaying or viewing a stereogram as defined in claim 14, comprising a further partitioning of MM into a range MS and a range MD, FS having a transmittance TSMS of MS and a transmittance TSMD of MD, and FD having a transmittance TDMS of MS and a transmittance TDMD of MD, such that: - TSMD is no greater than 6-. of TDMD; and
- TDMS is no greater than 6% of TSMS .
17. An apparatus for displaying a stereogram, the stereogram comprising a first part-image S and a second part-image D, the apparatus comprising a first optical filter FS and a second optical filter FD as defined any of claims 13-16, the apparatus comprising:
- means for transmitting a projection, or trans- mittmg light forming a projection, of at least a compo¬ nent CS of S through FS ;
- means for transmitting a projection, or transmitting light forming a projection, of at least a component CD of D through FD; and - means for adjusting the transmitted projections of S and D, or components CS and CD of S and D, to bring at least a first point PS m the transmitted projection of S or CS, into coalescence with a second point PD m the transmitted projection of D or CD.
18. An apparatus for displaying a stereogram, as defined m claim 17, the apparatus further comprising:
- means, such as electronic filtering means, for removing from the second part-image D or its electronic representation, a substantial part of its contribution after the first spectral distribution K; and
- means, such as electronic filtering means, for removing from the first part-image S or its electronic representation, a substantial part of its contribution after the second spectral distribution L, the distribu¬ tion L being excluded by the filter FS and the distribu¬ tion K being excluded by the filter FD .
19. An apparatus for displaying a stereogram, as defined m claim 17, the apparatus further comprising:
- a first image-forming unit USK, a second image-forming unit USM, a third image-forming unit UDM and a fourth image-forming unit UDL;
- means for partitioning from S a first contribution CSL and a second contribution CSM;
- means for partitioning from D a first contribution CDK and a second contribution CDM; - means for passing light emitted by USK through
FS;
- means for passing light emitted by USM through
FS;
- means for passing light emitted by UDM through FD; and
- means for passing light emitted by UDL through
FD.
20. A physically realized image, for example a printed or photographically reproduced image, the image comprising:
- a first component image PK, observable or re- recordable through a first colour separation filter FK;
- a second component image PL, observable or re- recordable through a second colour separation filter FL;
- a third component image PM, observable or re- recordable through a third colour separation filter FM, such that
- a disparity map constructed from a pair of im- ages, comprising as a first image a luminance level image
LIK made from PK and as a second image a luminance level image LIL made from PL, differs substantially from a constant-parallax map; - a first energy difference, being the difference between energies in the high-frequency range of the component image PM and in the high-frequency range of the component image PK being substantially different from zero; and
- a second energy difference, being the difference between energies in the high-frequency range of the component image PM and in the high-frequency range of the component image PL being substantially different from zero.
PCT/DK2000/000448 1999-08-10 2000-08-10 Methods and apparatuses for encoding and displaying stereograms WO2001011894A2 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
AU64275/00A AU6427500A (en) 1999-08-10 2000-08-10 Methods and apparatuses for encoding and displaying stereograms
EP00951279A EP1245121A2 (en) 1999-08-10 2000-08-10 Methods and apparatuses for encoding and displaying stereograms

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
DKPA199901115 1999-08-10
DKPA199901115 1999-08-10
DKPA200000748 2000-05-05
DKPA200000748 2000-05-05

Publications (3)

Publication Number Publication Date
WO2001011894A2 true WO2001011894A2 (en) 2001-02-15
WO2001011894A3 WO2001011894A3 (en) 2002-07-25
WO2001011894A8 WO2001011894A8 (en) 2004-05-21

Family

ID=26065256

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/DK2000/000448 WO2001011894A2 (en) 1999-08-10 2000-08-10 Methods and apparatuses for encoding and displaying stereograms

Country Status (3)

Country Link
EP (1) EP1245121A2 (en)
AU (1) AU6427500A (en)
WO (1) WO2001011894A2 (en)

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2366114A (en) * 2000-06-30 2002-02-27 Mark Thomas Dawson Anaglyphic production method using colour filtering.
US7046271B2 (en) 2000-01-25 2006-05-16 X3D Technologies Gmbh Method and system for the three-dimensional representation
WO2008155213A1 (en) * 2007-06-19 2008-12-24 Christophe Lanfranchi Method and equipment for producing and displaying stereoscopic images with coloured filters
WO2008140787A3 (en) * 2007-05-09 2009-01-29 Dolby Lab Licensing Corp System for 3d image projections and viewing
US7784938B2 (en) 2007-05-09 2010-08-31 Dolby Laboratories Licensing Corporation Method and system for shaped glasses and viewing 3D images
US7959295B2 (en) 2007-05-18 2011-06-14 Dolby Laboratories Licensing Corporation Spectral separation filters for 3D stereoscopic D-cinema presentation
DE102010024666A1 (en) * 2010-06-18 2011-12-22 Hella Kgaa Hueck & Co. Method for optical self-diagnosis of a camera system and apparatus for carrying out such a method
EP2605520A1 (en) * 2011-12-15 2013-06-19 Thomson Licensing Anaglyphic stereoscopic image capture device
EP2611172A1 (en) * 2011-12-28 2013-07-03 Thomson Licensing Method for generating an anaglyph image
WO2015090960A1 (en) * 2013-12-19 2015-06-25 Robert Bosch Gmbh Optical stereo system
US9313482B2 (en) 2006-04-05 2016-04-12 Bose Corporation Forming spectral filters
US10809543B2 (en) 2017-01-23 2020-10-20 Dolby Laboratories Licensing Corporation Glasses for spectral and 3D imaging
CN113674371A (en) * 2021-08-06 2021-11-19 吉林大学 Three-dimensional element image array coding method based on five-dimensional secondary kernel modeling

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4743965A (en) * 1985-05-07 1988-05-10 Nippon Hoso Kyokai Stereoscopic television picture transmission system
WO2000023845A2 (en) * 1998-10-20 2000-04-27 Soerensen Svend Erik Boerre A method for recording and viewing stereoscopic images in colour using multichrome filters

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4743965A (en) * 1985-05-07 1988-05-10 Nippon Hoso Kyokai Stereoscopic television picture transmission system
WO2000023845A2 (en) * 1998-10-20 2000-04-27 Soerensen Svend Erik Boerre A method for recording and viewing stereoscopic images in colour using multichrome filters

Cited By (35)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7046271B2 (en) 2000-01-25 2006-05-16 X3D Technologies Gmbh Method and system for the three-dimensional representation
GB2366114B (en) * 2000-06-30 2004-12-22 Mark Thomas Dawson Improvements in and relating to method and apparatus for producing anaglyphic 3-D images.
GB2366114A (en) * 2000-06-30 2002-02-27 Mark Thomas Dawson Anaglyphic production method using colour filtering.
US9313482B2 (en) 2006-04-05 2016-04-12 Bose Corporation Forming spectral filters
US9547179B2 (en) 2007-05-09 2017-01-17 Dolby Laboratories Licensing Corporation Method and system for shaped glasses and viewing 3D images
US9958693B2 (en) 2007-05-09 2018-05-01 Dolby Laboratories Licensing Corporation System for 3D image projections and viewing
US7784938B2 (en) 2007-05-09 2010-08-31 Dolby Laboratories Licensing Corporation Method and system for shaped glasses and viewing 3D images
US11585971B2 (en) 2007-05-09 2023-02-21 Dolby Laboratories Licensing Corporation System for 3D image projections and viewing
US10802293B2 (en) 2007-05-09 2020-10-13 Dolby Laboratories Licensing Corporation System for 3D image projections and viewing
KR102024512B1 (en) 2007-05-09 2019-09-24 돌비 레버러토리즈 라이쎈싱 코오포레이션 System for 3d image projections and viewing
US10338399B2 (en) 2007-05-09 2019-07-02 Dolby Laboratories Licensing Coporation Method and system for shaped glasses and viewing 3D images
CN102998802A (en) * 2007-05-09 2013-03-27 杜比实验室特许公司 System for 3D image prjections and viewing
US8459796B2 (en) 2007-05-09 2013-06-11 Dolby Laboratories Licensing Corporation Method and system for shaped glasses and viewing 3D images
KR20180108882A (en) * 2007-05-09 2018-10-04 돌비 레버러토리즈 라이쎈싱 코오포레이션 System for 3d image projections and viewing
WO2008140787A3 (en) * 2007-05-09 2009-01-29 Dolby Lab Licensing Corp System for 3d image projections and viewing
US9921412B2 (en) 2007-05-09 2018-03-20 Dolby Laboratories Licensing Corporation Method and system for shaped glasses and viewing 3D images
US8503078B2 (en) 2007-05-09 2013-08-06 Dolby Laboratories Licensing Corporation Method and system for shaped glasses and viewing 3D images
US8537463B2 (en) 2007-05-09 2013-09-17 Dolby Laboratories Licensing Corporation Method and system for shaped glasses and viewing 3D images
CN102998802B (en) * 2007-05-09 2016-07-06 杜比实验室特许公司 3D display system and 3D watch glasses
US9146402B2 (en) 2007-05-09 2015-09-29 Dolby Laboratories Licensing Corporation Method and system for shaped glasses and viewing 3D images
US8403489B2 (en) 2007-05-18 2013-03-26 Dolby Laboratories Licensing Corporation Spectral separation filters for 3D stereoscopic D-cinema presentation
US7959295B2 (en) 2007-05-18 2011-06-14 Dolby Laboratories Licensing Corporation Spectral separation filters for 3D stereoscopic D-cinema presentation
FR2917845A1 (en) * 2007-06-19 2008-12-26 Christophe Brossier METHOD OF VISUALIZING A SEQUENCE OF IMAGES PRODUCING A RELIEF SENSATION
CN101755236B (en) * 2007-06-19 2012-07-04 特瑞维兹公司 Method and equipment for producing and displaying stereoscopic images with coloured filters
WO2008155213A1 (en) * 2007-06-19 2008-12-24 Christophe Lanfranchi Method and equipment for producing and displaying stereoscopic images with coloured filters
US8854431B2 (en) 2010-06-18 2014-10-07 Hella Kgaa Hueck & Co. Optical self-diagnosis of a stereoscopic camera system
DE102010024666A1 (en) * 2010-06-18 2011-12-22 Hella Kgaa Hueck & Co. Method for optical self-diagnosis of a camera system and apparatus for carrying out such a method
CN103163718A (en) * 2011-12-15 2013-06-19 汤姆森特许公司 Anaglyphic stereoscopic image capture device
EP2605520A1 (en) * 2011-12-15 2013-06-19 Thomson Licensing Anaglyphic stereoscopic image capture device
US9247233B2 (en) 2011-12-15 2016-01-26 Thomson Licensing Anagylphic stereoscopic image capture device
EP2611172A1 (en) * 2011-12-28 2013-07-03 Thomson Licensing Method for generating an anaglyph image
WO2015090960A1 (en) * 2013-12-19 2015-06-25 Robert Bosch Gmbh Optical stereo system
US10809543B2 (en) 2017-01-23 2020-10-20 Dolby Laboratories Licensing Corporation Glasses for spectral and 3D imaging
CN113674371A (en) * 2021-08-06 2021-11-19 吉林大学 Three-dimensional element image array coding method based on five-dimensional secondary kernel modeling
CN113674371B (en) * 2021-08-06 2023-11-03 吉林大学 Three-dimensional element image array coding method based on five-dimensional quadratic kernel modeling

Also Published As

Publication number Publication date
EP1245121A2 (en) 2002-10-02
AU6427500A (en) 2001-03-05
WO2001011894A3 (en) 2002-07-25
WO2001011894A8 (en) 2004-05-21

Similar Documents

Publication Publication Date Title
EP1131658B1 (en) A method for recording and viewing stereoscopic images in colour using multichrome filters
US8723932B2 (en) Display of generalized anaglyphs without retinal rivalry
McKenna et al. Three dimensional visual display systems for virtual environments
US20020021832A1 (en) Method and apparatus for producing anaglyphic 3-D images
RU1836648C (en) Spectacles producing stereo image effect
EP0716330A2 (en) Apparatus for displaying image recognized by observer as stereoscopic image, and image pick-up apparatus
EP2544459A1 (en) Stereoscopic video display device and operation method of stereoscopic video display device
WO2001011894A2 (en) Methods and apparatuses for encoding and displaying stereograms
CA2768701A1 (en) Method for crosstalk correction for three-dimensional (3d) projection
JP2010531102A5 (en)
US20130033713A1 (en) Apparatus and method of forming image, terminal and method of print control, and computer-readable medium
CN102549476B (en) Method and system for color correction for three-dimensional (3D) projection
Woods et al. Characterizing and reducing crosstalk in printed anaglyph stereoscopic 3D images
JP2001298754A (en) Crosstalk cancellation method and system for stereoscopic image
Zeng et al. Printing anaglyph maps optimized for display
EP0107092A2 (en) Method and apparatus for generating multicolor and/or threedimensional pictures
US3221600A (en) Method for forming and viewing color separation stereoscopic images
Safonov et al. Anaglyph Printing
US8363976B2 (en) Method of rendering cross images unobservable particularly in projected or printed photographic reproductions, and reproduction thus obtained
KR20130018517A (en) Apparatus and method for forming image, terminal and method for print control, and computer-readable recording medium
Dequal et al. A new support for teaching and research in photogrammetry: The stereoscopic classroom
JPS62125341A (en) Stereoscopic video photographing method
Dunkley New 3D from 2D visual display process
BE571553A (en)
WO2013023322A1 (en) Multimodal color filtering three-dimensional image display method

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A2

Designated state(s): AE AG AL AM AT AT AU AZ BA BB BG BR BY BZ CA CH CN CR CU CZ CZ DE DE DK DK DM DZ EE EE ES FI FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ PL PT RO RU SD SE SG SI SK SK SL TJ TM TR TT TZ UA UG US UZ VN YU ZA ZW

AL Designated countries for regional patents

Kind code of ref document: A2

Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE BF BJ CF CG CI CM GA GN GW ML MR NE SN TD TG

121 Ep: the epo has been informed by wipo that ep was designated in this application
DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
WWE Wipo information: entry into national phase

Ref document number: 2000951279

Country of ref document: EP

WWE Wipo information: entry into national phase

Ref document number: 10069041

Country of ref document: US

REG Reference to national code

Ref country code: DE

Ref legal event code: 8642

AK Designated states

Kind code of ref document: A3

Designated state(s): AE AG AL AM AT AT AU AZ BA BB BG BR BY BZ CA CH CN CR CU CZ CZ DE DE DK DK DM DZ EE EE ES FI FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ PL PT RO RU SD SE SG SI SK SK SL TJ TM TR TT TZ UA UG US UZ VN YU ZA ZW

AL Designated countries for regional patents

Kind code of ref document: A3

Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE BF BJ CF CG CI CM GA GN GW ML MR NE SN TD TG

WWP Wipo information: published in national office

Ref document number: 2000951279

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: JP

CFP Corrected version of a pamphlet front page
CR1 Correction of entry in section i

Free format text: IN PCT GAZETTE 07/2001 DUE TO A TECHNICAL PROBLEMAT THE TIME OF INTERNATIONAL PUBLICATION, SOME INFORMATION WAS MISSING UNDER (81). THE MISSING INFORMATION NOW APPEARS IN THE CORRECTED VERSION

WWW Wipo information: withdrawn in national office

Ref document number: 2000951279

Country of ref document: EP