`Geshwind et al.
`
`4,925,294
`[11] Patent Number:
`[45] Date of Patent: May 15, 1990
`
`‘[54] METHOD TO CONVERT TWO
`DIIVIENSIONAL MOTION PICTURES FOR
`THREE-DIMENSIONAL SYSTEMS
`
`76 In
`1
`[
`]
`ventors
`
`-
`
`-
`
`.
`
`1 4_ 4 -
`811 4132Mldland
`Anthony H. Handal, Blue Chip La‘,
`Westport, Conn. 06880
`
`[21] Appl. No.:
`[22] Filed:
`
`Dec. 17, 1986
`
`[51] Int. Cl.5 ............................................ .. G03B 35/14
`[52] US. Cl. ...................................... .. 352/57; 352/86;
`355/40
`[58] Field of Search ............. .. 355/40, 52, 77; 352/43,
`352/57, 85, 86, 129; 358/88, 89, 22, 81, 160
`
`[56]
`
`References Cited
`
`_ U's' PATENT DOCUMENTS
`3,772,465 11/1973 Vlahos et a1. ....................... .. 355/40
`3,824,336 7/1974 Gould et a1. .
`355/52
`4,606,625 8/1986 Geswind ..... ..
`352/85 X
`4,809,065 2/1989 Harris et a1. ........................ .. 358/88
`Primary Examiner-L. T. Hix
`Assistant Examiner-D. Rutledge
`[57]
`ABSTRACI,
`The present invention relates to the computer-assisted
`processing of standard two-dimensional motion pictures
`to generate processed image sequences which exhibit
`some three-dimensional depth effects when viewed
`under appropriate conditions.
`
`44 Claims, 2 Drawing Sheets
`
`2-0~ FILH OR
`VIDEO INPUT
`1o
`
`/
`\
`
`I
`1
`F a l SCANNER | orsrunr
`R u
`1n |
`q2
`MIL“- -----
`n F I
`HEHDRY
`E E
`In:
`R l--—-"--——-
`l COMPUTER INTERFACE
`no:
`115
`
`CRT VIDEO
`DISPLAY
`2o
`
`E
`
`PEN as
`
`V
`
`COHPUTERICPU
`so
`
`TABLET
`30
`
`/ KEYBOARD
`55
`
`oumn
`INTERFACE
`60
`
`3-0 VTR
`7o
`
`I
`1
`.
`|
`5-0 run
`L) RECORDER
`7s
`
`Legend3D, Inc. Ex. 2022-0001
`IPR2016-01243
`
`
`
`US. Patent May 15, 1990
`
`Sheet 1 of 2
`
`4,925,294
`
`LID
`
`FIGURE 1
`
`Legend3D, Inc. Ex. 2022-0002
`IPR2016-01243
`
`
`
`US. Patent May 15,1990
`
`Sheet 2 0f'2
`
`4,925,294 ,
`
`DD
`2v
`_I
`
`E
`
`m m w s _ w w MW IILIIPIII
`
`HI BUFFER 0
`F0 FRAME U:
`
`E u- _ M _ E E
`A _ P P
`
`D _ Y H. _ T U
`
`C _ M M
`
`/\ .l|l.l..l_un _ A I
`
`\ _mq m w
`R41 E R R
`M _m V m
`
`E
`
`0 m N
`
`DVI E
`
`V L E 3 P L TS B RI A CD T
`IA P T0
`
`5 3
`
`> _
`
`Au. _Cu.
`Y2_ _Es <
`
`H _ m
`m _ _ m
`
`50
`
`KEYBOARD
`55
`
`OUTPUT
`INTERFACE
`60
`
`3-D VTR
`70
`
`l-»
`
`3-D FILM
`RECORDER
`75
`
`FIGURE 2
`
`Legend3D, Inc. Ex. 2022-0003
`IPR2016-01243
`
`
`
`1
`
`4,925,294
`
`METHOD TO CONVERT TWO DIIVIENSIONAL
`MOTION PICTURES FOR THREE-DIMENSIONAL
`SYSTEMS
`
`TECHNICAL FIELD
`The invention relates to a method for converting
`existing ?lm or videotape motion pictures to a form that
`can be used with three-dimensional systems for broad
`cast or exhibition.
`
`10
`
`2
`sion receiver that can be viewed as a normal 2-D picture
`without glasses. Very inexpensive glasses, with one
`light and one dark lens, accentuate the differential pro
`cessing of the image, as viewed by each eye, to produce
`a 3-D depth effect.
`Practical, inexpensive, compatible (with standard
`TV) 3-D‘ television may now become widespread. In
`addition to materials speci?cally produced for the new
`system (or other 3-D systems) conversion of standard
`2-D programs to 3-D format would provide additional
`product to broadcast using the new compatible system
`(or for other 3-D projection systems).
`
`BACKGROUND ART
`With the advent of stereophonic sound, various tech
`niques were developed to convert or ‘re-process’ exist
`ing monophonic programs for stereophonic broadcast
`or recording systems. These included modifying the
`equalization phase or tonal qualities of separate copies
`of the monophonic program for the left and right chan
`nels. While true stereophonic or binaural effects may
`20
`not have been achieved, the effects were much im
`proved over feeding the identical monophonic signal to
`both channels.
`Similarly, with the almost universal use of color pro
`duction, exhibition and broadcast systems for motion
`pictures and television, systems have been developed to
`convert existing monochrome or black and white mate
`rials to color programs. Such a systems is described in _
`applicant Geshwind’s Pat. No. 4,606,625, issued Aug.
`19, 1986. The results of these colorized products, while
`not always identical to true color motion pictures, are
`more suitable than black and white for color systems.
`There have been a number of systems for exhibition
`or display of left- and right-eye pairs of binocular mo
`tion pictures. Early systems required two completely
`redundant projection or display systems; e.g. two ?lm
`projectors or CRT television displays, each routed to
`one eye via mirrors. Other systems require either com
`plicated and expensive projection or display systems, or
`expensive ‘glasses’ to deliver two separate images. For
`example:
`red- and green-tinted monochrome images are both
`projected or displayed to be viewed through
`glasses with left and right lenses tinted either red or
`green;
`two full-color images are projected through mutually
`perpendicular polarized ?lters and viewed through
`glasses with lenses that are also polarized in the
`same manner;
`.
`left and right images are displayed on alternate odd
`and even ?elds (or frames) of a standard (or high
`scan rate) television CRT and are viewed through
`‘glasses’ with shutters (either rotating blades or
`?ickering LCDs, for example) that alternate the
`view of left and right eyes in synchrony with the
`odd or even ?elds of the CRT.
`Of the above systems, the second is not at all usable
`with standard home television receivers, the third re
`quires very expensive ‘glasses’ and may ?icker with
`standard home receivers, and the ?rst produces only
`strangely tinted monochrome images. Further, none of 60
`the systems may be broadcast over standard television
`for unimpeded viewing without special glasses.
`Thus, until now, compatible (i.e., viewable as two-di
`mensional, without glasses) home reception of b. 3-D
`images was not possible. However, a new system,
`which takes advantage of differential processing of left
`and right-eye images in the human perceptual system,
`delivers a composite image on a standard home televi
`
`SUMMARY OF THE INVENTION
`Applicant Handal’s previous application, No.
`479,679, ?led Mar. 28, 1983 and now abandoned, relates
`a process and apparatus for deriving left- and right-eye
`pairs of binocular images, from certain types of two
`dimensional ?lm materials. In particular, the materials
`must consist of separate foreground and background
`elements, such as cartoon animation cells and back
`ground art. By taking into account the parallax between
`scenes as viewed by the left and right eyes, two images
`are prepared where foreground elements are shifted
`with respect to background elements by an amount that
`indicates their depth in the third dimension. Two-di
`mensional motion pictures that consist of a series of
`single composite images could not be converted to
`three-dimensional format, by this technique, without
`?rst being separated into various background and fore
`ground elements.
`Once committed to two~dimensional motion pictures,
`the separation and depth information for various scene
`elements, in the third dimension, are lost. Thus, the
`separation of two-dimensional image sequences into
`individual image elements and the generation of three
`dimensional depth information for each such image
`element are not simple or trivial tasks, and are the fur
`ther subject of the instant invention.
`In accordance with the invention, standard two-di
`mensional motion picture ?lm or videotape may be
`converted or processed, for use with three-dimensional
`exhibition or transmission systems, so as to exhibit at
`least some three—dimensional or depth characteristics.
`Separation of a single 2-D image stream into diverse
`elements is accomplished by a computer assisted,
`human operated system. (If working from discrete 2-D
`?lm sub-components, such as animation elements, the
`separation step may be omitted.) Depth information is
`assigned to various elements by a combination of human
`decisions and/or computer analyses and resulting im
`ages of three-dimensional format are produced under
`computer control.
`
`BRIEF DESCRIPTION OF DRAWINGS
`A method for carrying out the invention is described
`in the accompanying drawings in which: ‘
`FIG. 1 is a diagram illustrating the relationship of a
`2-D source ?lm image, left and right image pairs, and a
`composite 3-D image frame.
`FIG. 2 is a schematic diagram of a system for carry
`ing out an implementation of the present invention.
`
`25
`
`30
`
`45
`
`55
`
`65
`
`DETAILED DESCRIPTION
`’
`An immense amount of standard, 2-D motion picture
`material exists in the form of ?lm and videotape. In -
`addition, certain materials exist in the form of discrete
`
`Legend3D, Inc. Ex. 2022-0004
`IPR2016-01243
`
`
`
`15
`
`20
`
`35
`
`4,925,294
`3
`4
`2-D image sub-components, such as animated cell and
`to objects as they get closer. Alternately, attention may
`background paintings. As the use of 3-D exhibition and
`be centered in the mid-ground with no parallax offset to
`broadcast systems becomes more widespread, the con
`mid-range objects, some parallax offset to close-range
`version of existing 2-D programs to a format that will
`objects and reverse parallax offset to far-range objects.
`The particular placement of objects and attention point
`exhibit at least some 3-D or depth effects, when used
`with 3-D systems, is desired.
`in the 3-D scene is as much an art as a science and is
`Extracting individual image elements or 3-D depth
`critical to the enjoyment of 3-D programs and, in any
`information from a 2-D film frame, or synthesizing 3-D
`event, is not meant to be the subject of this invention.
`information for those elements, entirely by computer
`FIG. 2 shows a schematic of a system to implement
`equipped with arti?cial intelligence, is not now practi
`the instant invention. A 2-D ?lm or video image 10 is
`cal. Therefore, the embodiment of the invention as
`input to a video monitor 20 and to the scanning portion
`described herein employs a high degree of human inter
`41 of frame buffer 40. Video monitor 20 is capable of
`action with the computer. However, as arti?cial intelli
`displaying either the 2-D image being input 10 or the
`gence progresses, a predominantly or completely auto
`output from display portion 42 of frame buffer 40.
`mated system may become practical and is within the
`Frame buffer 40 consists of an image scanner 41
`intended scope of the invention.
`which can convert the input image into digital form to
`FIG. 1 shows a frame 20 of a standard 2-D ?lm of
`be stored in a portion of frame buffer memory section
`video motion picture, consisting of a cross-hatched
`44, a display section 42 which creates a video image
`background plane 21, a large white circle 22 in the
`from the contents of a portion of memory section 44,
`mid-ground, and a small black square 23 in the fore
`and a computer interface section 43 which allows the
`ground. It is a 2-D representation of original 3-D scene
`computer CPU 50 to read from and write to the mem
`10 comprising elements 11, 12 and 13, which is not
`ory section 44.
`available for direct 3-D photography. After human
`Graphic input tablet 30 and stylus 35 allow the human
`identi?cation of individual image elements and depth
`operator to input position information to the computer
`assignment, the computer generates, from frame 10, a
`50 which can indicate the outline of individual image
`25
`pair of coordinated left 30 and right 40 images with
`elements, choice of a specific image element or part of
`backgrounds 31 and 41, circles 32 and 42, and squares 33
`an image element, depth speci?cation, choice of one of
`and 43 respectively. Note that the relative positions of '
`a number of functions offered on a ‘menu’, or other
`the square and circle are different in the left and right
`information. An image cursor can be displayed on mon
`images; this situation is similar to the parallax that might
`itor 20 by frame buffer 40 to visually indicate the loca
`result between left- and right-eye views if one were to
`tion or status of the input from the tablet 30 and pen 35.
`have viewed the original scene 10 directly. Three-di
`Text and numeric information can also be input by the
`mensional format frame 50 is generated by encoding the
`operator on keyboard 55.
`information in the left 30 and right 40 image pair, in a
`Computer CPU 50 is equipped with software which
`manner consistent with any one of a number of existing
`allows it to interpret the commands and input from the
`3-D systems. The speci?c operation of these various
`human operator, and to process the digitized 2-D infor
`3-D systems is not the subject of the instant invention.
`mation input from 2-D frame 10 into digital 3-D frame
`Alternately, the steps of generating and encoding 3-D
`information, based on said human commands and input.
`information may be combined such that 3-D format
`Said digital 3-D frame information is then output by
`frame 50 may be processed directly from 2-D frame 20
`output interface 60 (which may be similar to frame
`without generating left 30 and right 40 image pairs. In
`buffer 40 or of some other design) to a videotape re
`either case, 3-D format frame 50 when viewed by
`corder 70 or to a ?lm recorder 75, capable of recording
`human 60 through 3-D glasses 70 is perceived as 3-D
`3-D format frames.
`scene 80, containing elements 81, 82 and 83, which has
`The system as described above operates as follows. A
`at least some of the 3-D characteristics of original 3-D
`frame of the 2-D program is displayed on the video
`scene 10.
`monitor for viewing by the human operator. With the
`Various systems for .the encoding, display, projec
`tablet and stylus the operator outlines various 2-D
`tion, recording, transmission or viewing of 3-D images
`image areas to indicate to the computer the boundaries
`exist, and new systems may be developed. Speci?cally,
`of various image elements to be separately processed.
`various techniques for specifying, encoding and view
`(For materials, such as animation components, that
`ing 3-D information may now, or come to, exist, which
`already exist as separate elements, the previous stage of
`do not make use of parallax offset and/or left and right
`the process may be skipped.) Depth position informa
`image pairs and/or viewing glasses, or which embody
`tion, in the third dimension, is determined for each
`new techniques or changes and improvements to cur
`image element, by a combination of operator input and
`rent systems. Further, such systems may integrate infor
`computer analysis. Left and right image pairs or a 3-D
`mation from more than one 2-D source frame 20 into a
`composite image is processedby the computer, from the
`single resultant 3-D frame 50. The speci?cs of operation
`2-D input image, based on the computer software and
`of such systems is not the subject of the instant inven
`operator instructions. Depending upon the particular
`tion, however, preparation of 2-D program material for
`3-D system to be used, left- and right-image pairs may
`such systems is.
`or may not be the ?nal stage or an intermediate stage or
`The offsets shown for elements 31, 32 and 33 in left
`bypassed entirely. Further, for some 3-D systems, infor
`frame 30 and elements 41, 42 and 43 in right frame 40
`mation from more than one 2-D source frame may be
`are meant to be illustrative and do not necessarily fol
`combined into one 3-D frame (or frame pair). In any
`low the correct rules for image parallax. In fact, de
`event, the ?nal 3-D information is then collected on a
`pending upon where viewer attention is meant to be
`videotape or film recorder. The process is then repeated
`centered, different rules may apply. For example, one
`for additional 2-D frames.
`technique is to give no parallax offset to far background
`Each image element may be given a uniform depth
`elements and to give progressively more parallax offset
`designation which may cause the perception of ‘card
`
`30
`
`45
`
`50
`
`55
`
`60
`
`65
`
`Legend3D, Inc. Ex. 2022-0005
`IPR2016-01243
`
`
`
`15
`
`20
`
`4,925,294
`5
`6
`board cut-out’ characters. Alternately, different por
`late. However, there are well known techniques, that
`tions of a single image element may be given different
`have been developed for computer image synthesis, that
`depth designations with the computer interpolating
`allow for the sorting, intersection and display of di
`depth coordinates over the entire element. For example,
`verse, overlapping 3-D image elements.
`an image element positioned diagonally in a frame may
`As image elements are separated from background
`have its right edge designated to be closer than its left
`scenes or each other, holes may develop. The computer
`software will compensate for this by filling in missing
`edge. Alternately, one feature of an image element, say
`a person’s nose in a close-up, might be designated as
`information in one particular 2-D frame with the equiv
`being closer than the rest of the image element. In such
`alent part of the image element from earlier or later
`frames whenever possible. Alternately, material to ‘plug
`manner, the computer would be instructed to interpo
`late and process depth information over the entire
`holes’ may be created by the operator or by the com
`image element. Such processing of the image element,
`puter from adjacent areas or may be newly synthesized.
`in accordance with varying depth information, may
`Further, as two image elements are separated from each
`result in the stretching, skewing or other distortion of
`other, the forwardmost of the two may be moved closer
`the image element.
`to the viewer; hence, it may be appropriate to make it
`Depth interpolation may also be carried out over
`larger. Thus, the increased size may cover the gaps in
`time, between frames. For example, an image element
`the rearmost element of the two.
`might be designated to be close in one frame and to be
`As part of the processing to be performed on the 2-D
`far away in a later frame. The depth of the image ele
`source image, individual image elements or composite
`ment may be interpolated for the frames between the
`images, additional effects may be programmed into the
`two designated frames. Further, the position of the
`computer to heighten the sense of depth. For example,
`image element in the two dimensions of the film frame,
`shadows cast by one image element on another element
`and the steps of the outline separating the image ele
`or the background may be calculated and inserted; far
`ment from the rest of the source image, may also be
`distant landscape elements may be made hazy or bluish
`interpolated over time, between frames Linear and non
`to indicate remoteness; different image elements may be
`25
`linear interpolation techniques are well known and may
`blurred or sharpened to simulate depth-of-?eld; or, the
`be readily applied.
`distance between close objects may be exaggerated to
`It is also possible to add random noise to the depth '
`emphasize their separation. There are, of course, other
`information to eliminate the appearance of ?at objects
`technical or artistic techniques that can be used to indi
`moving in space and to help achieve greater realism.
`cate depth in an image and which may also be incorpo
`30
`For a particular image element, depth position may
`rated into the image processing programs and would
`be derived by the computer, alone or with operator
`therefore be part of the invention as described herein.
`input, by measuring the size of the image element in a '
`Therefore, the above examples are illustrative and
`frame. For example, once a person were outlined, his
`should not be construed as limiting the scope of the
`overall height in the 2-D film frame might be extracted
`invention. Alternately, depth information may be inten
`by the computer as an indication of depth distance.
`tionally distorted for effect or for artistic purposes.
`using knowledge, or making assumptions, about the
`Improvements may be made to the ?nal product by
`object’s real size, and the characteristics of the camera
`including new image elements that were not part of the
`and lens, the distance of the object from the camera may
`original 2-D source image. These could include 2-D
`be calculated by applying well known principles of
`image elements that are then assigned depth values, 3-D
`40
`perspective and optics. Alternately, if overall size can
`image elements created by 3-D photography and then
`not be easily extracted by the computer, key points
`entered into the computer as left- and right-image pairs,
`might be indicated by the operator for the computer to
`for example, or synthetic 3-D computer generated
`graphics. In particular, since computer generated image
`measure. Comparing the change of size of an image
`element in several frames will provide information
`elements can be created with depth information, they
`45
`about the movement of the object in the third dimen
`can be easily integrated into the overall 3-D scene with
`vivid effect. For example, a 3-D laser blast could be
`son.
`created by computer image synthesis such that it would
`It should be noted that horizontal parallax offsets are,
`in turn obscure'and be obscured by other image ele
`by far, more obvious, due to the fact that our eyes are
`separated in the horizontal direction. However, vertical
`ments in an appropriate manner and might even be
`50
`offset differences between left- and right-eye views may
`created so as to appear to continue beyond the front of
`also be appropriate.
`the screen into ‘viewer space’.
`As various image elements are separated and assigned
`Animated ?lm components usually have relatively
`depth values, a situation develops where diverse objects
`few background paintings, which are kept separate
`exist in a ‘three-dimensional space’ within the computer.
`from the animated characters in the foreground. For
`It should be noted that, in order to display a realistic
`these, or for live 2-D filmed scenes, once the fore
`representation of the entire scene, forwardmost objects
`ground elements have been separated from the’ back
`ground, the flat 2-D backgrounds may be ‘replaced by
`must obscure all or part of rearmost objects with which
`3-D backgrounds. The 3-D backgrounds might consist
`they overlap (except in the case where a forwardmost
`object were transparent). When generating left- and
`of computer generated graphics, in which case depth
`right-eye views, the pattern of overlap of image ele
`information for the various elements of the background
`ments, and thus the pattern of obscuring of image ele
`would be available at the time of the background cre
`ments will, in general, be different. Further, for image
`ation. Alternatively, 3-D backgrounds might be created
`elements that have been assigned non-uniform depth
`using 3-D photography, in which case depth informa
`values (e.g., image elements that are not flat or not in a
`tion for the background elements may be derived, by
`plane perpendicular to the third dimension) the inter
`the computer, from the comparison of the left- and
`right-image pairs of the 3-D background photographs
`section between these image elements, for the purpose
`and applying known image processing and pattern rec
`of one obscuring the other, may be non-trivial to calcu
`
`35
`
`65
`
`Legend3D, Inc. Ex. 2022-0006
`IPR2016-01243
`
`
`
`LII
`
`15
`
`20
`
`25
`
`ments.
`
`'
`
`4,925,294
`7
`8
`ognition techniques. Alternately, depth information to
`4. A method as in claim 2 comprising the additional
`create 3-D backgrounds may be speci?ed otherwise by
`step of:
`operator input and/or computer processing. Once
`g. encoding said left and right processed image pair
`for viewing, by coloring each of said pair different
`depth information is available for the various back~
`ground elements, it may compared to depth information
`colors.
`for the other image elements with which it is to be
`5. A method as in claim 2 comprising the additional
`combined in each frame. Thus, the 3-D background
`step of: ‘
`g. encoding said left and right processed image pair
`may itself live some depth and, in effect, be a ‘set’ within
`for viewing, by passing each of said pair through
`which other image elements may be positioned in front
`mutually perpendicularly polarized ?lters.
`of, behind or intersecting various background elements
`6. A method as in claim 2 comprising the additional
`for added realism.
`step of:
`It should be noted that, intended purpose and soft
`g. encoding said left and right processed image pair
`ware algorithms aside, the design and operation of the
`for viewing, by incorporating said image pair into a
`3-D conversion system described herein has many simi
`video signal suitable for displaying each of said pair
`larities with the black and white footage colorization
`alternately on a video display.
`system described in applicant Geshwind’s Pat. No.
`7. As method as in claim 1, wherein said step f results
`4,606,625. They are both computer-aided systems
`in a processed image frame such that, when viewed
`which display standard ?lm images to an operator,
`through glasses with one dark and one light lens, 3
`allow the operator to input information separating vari
`dimensional effects are perceived.
`ous image elements within the frame, allow the operator
`8. A method as in claim 1 comprising the additional
`to specify attributes (color in one case, depth in the
`step of:
`other) for the image elements, and cause the computer
`g. recording said processed image frame.
`to process new image frames from the original, based on
`9. A method as in claim 7 comprising the additional
`the operator input. Thus, the processing of 2~D black
`step of:
`and white footage to add both color and depth informa
`g. recording said processed image frame.
`tion would be much more ef?cient than implementing
`10. A method as in claim 1 wherein said steps are
`each process separately.
`applied to successive frames in a motion picture se
`The scope of the instant invention is the conversion
`quence.
`or processing of 2-D program material for use with any
`11. A method as in claim 2 wherein said steps are
`3-D exhibition or distribution system now in use or later
`applied to successive frames in a motion picture se
`developed, but not the speci?c method of operation of
`quence.
`.
`any particular 3-D system.
`12. A product produced by the method described in
`It will thus be seen that the objects set forth above,
`claim 7.
`among those made apparent from the proceeding de
`13. A product produced by the method described in
`scription, are efficiently attained and certain changes
`claim 10.
`may be made in carrying out the above method and in
`14. A product produced by the method described in
`the construction set forth. Accordingly, it is intended
`claim 11.
`that all matter contained in the above description or
`15. A product produced by the method described in
`shown in the accompanying ?gures shall be interpreted
`claim 1.
`as illustrative and not in a limiting sense.
`16. A method as in claim 1 wherein at least one of said
`Now that the invention has been described, what is
`processed image elements produced in step e is a
`claimed as new and desired to be secured by Letters
`shadow element.
`Patent is:
`17. A method as in claim 1 wherein at least one of said
`1. A method of converting a two-dimensional image
`image elements in step e is processed to include addi
`frame into a three-dimensional image frame consisting
`tional two-dimensional image information not con
`of the steps of:
`tained in the original unprocessed two-dimensional im
`a. inputing a frame of a two-dimensional image into a
`age.
`computer;
`18. A method as in claim 17 wherein said additional
`b. specifying at least two individual image elements in
`two-dimensional image information is derived from
`the two-dimensional image;
`another image.
`c. separating the two-dimensional image into said
`19. A method as in claim 1 wherein said processed
`image elements;
`image elements in step f are combined with at least one
`d. specifying three-dimensional information for at
`additional 3-D image element not derived from the
`least one of said image elements;
`source image to create said processed image frame.
`e. processing at least one of said image elements to
`20. A method as in claim 19 wherein said additional
`incorporate said three-dimensional information and
`3-D image element is derived from a 3A-D photograph.
`create at least one processed image element;
`21. A method as in claim 19 wherein said additional
`f. generating at least one processed image frame com
`3-D image element is derived from a computer gener
`prising at least'one of said processed image ele
`ated 3-D image.
`22. A method as in claim 1 wherein said three-dimen
`sional information for at least one of said image ele
`ments in step d is speci?ed only at certain points and is
`interpolatively derived for other points on said image
`element.
`23. A method as in claim 10 wherein said three-di
`mensional information for at least one of said image
`elements in step d is speci?ed only for certain frames
`
`2. A method as in claim 1 wherein said step f results
`in the generation of a left and right pair of processed
`image frames.
`'
`3. A method as in claim 2 comprising the additional
`step of:
`g. combining said left and right image pair into a
`single processed image.
`
`35
`
`45
`
`50
`
`55
`
`65
`
`Legend3D, Inc. Ex. 2022-0007
`IPR2016-01243
`
`
`
`4,925,294
`10
`and is temporally interpolated for frames between said
`d. a means for specifying three-dimensional informa
`tion for said individual image elements;
`certain frames.
`e. a means for processing said individual image ele
`24. A method as in claim 10 wherein said speci?cation
`ments to create processed image elements;
`of at least one of said image elements in step b is speci
`f. a means for combining said processed image ele
`?ed only for certain frames and is temporally interpo
`ments into a processed image sequence;
`lated for frames between said certain frames.
`g. a means for outputing said three-dimensional image
`25. A method as in claim 1 wherein random noise is
`added to the three-dimensional information speci?ed in
`sequence;
`'
`h. a means for recording said three-dimensional image
`step d.
`sequence.
`26. A method as in claim 1 wherein at least some of
`37. A method of converting a two-dimensional image
`said three-dimensional information speci?ed in step d is
`frame into a three-dimensional image frame consisting
`derived from the measurement of at least one aspect of
`of the steps of:
`an image element.
`a. inputing frames of a two-dimensional image se
`27. A method as in claim 26 wherein said aspect per
`quence into a computer;
`tains to geometry.
`b. specifying at least two individual image elements in
`28. A method as in claim 26 wherein said aspect per
`the two-dimensional image sequence;
`tains to illumination.
`0. separating the two-dimensional images into said
`29. A method as in claim 10 wherein at least some of
`image elements;
`said three~dimensional information speci?ed in step dis
`d.
`20
`' specifying three-dimensional information for at
`derived from the measurement of the change of at least
`least one of said image elements;
`one aspect of an image element in successive frames.
`e. processing at least one of said image elements to
`30. A method as in claim 29 wherein said aspect per
`incorporate said three-dimensional information and
`tains to geometry.
`create at least one processed image elements;
`31. A method as in claim 29 wherein said aspect per
`f. generating a sequence of processed image frames
`tains to illumination.
`comprising at least one of said processed image
`32. A method as in claim 1 wherein said two-dimen
`elements said generation to be of such a nature so as
`sional image frame is a black and white image frame, '
`to exhibit three-dimensional depth characteristics
`said image elements are black and white image ele
`when viewed using a complementary display sys
`ments, and said processing includes the process of add
`tem.
`ing color to at least one of said black and white image
`38. A method as in claim 37 comprising the additional
`step:
`elements.
`g. transmission of said three-dimensional image se
`33. A method as in claim 32 wherein said steps are
`quence.
`applied to successive frames in a motion picture se
`39. A method as in claim 33 comprising the additional
`quence.
`step:
`34. A product produced by the method of claim 33.
`g. recording of said three-dimensional image se
`35. An apparatus for converting a two-dimensional
`quence.
`image frame into a three-dimensional image frame com
`40. A method as in claim 37 wherein said two