+

WO1997015150A1 - Procede et dispositif de formation d'images en trois dimensions - Google Patents

Procede et dispositif de formation d'images en trois dimensions Download PDF

Info

Publication number
WO1997015150A1
WO1997015150A1 PCT/JP1996/003027 JP9603027W WO9715150A1 WO 1997015150 A1 WO1997015150 A1 WO 1997015150A1 JP 9603027 W JP9603027 W JP 9603027W WO 9715150 A1 WO9715150 A1 WO 9715150A1
Authority
WO
WIPO (PCT)
Prior art keywords
video signal
eye
image
spatial
converted
Prior art date
Application number
PCT/JP1996/003027
Other languages
English (en)
French (fr)
Inventor
Seisuke Oba
Takao Kumagai
Hitoshi Fukudome
Nobuyuki Minami
Nariyasu Hamahata
Original Assignee
Sony Corporation
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 Sony Corporation filed Critical Sony Corporation
Priority to GB9709228A priority Critical patent/GB2309847B/en
Priority to US08/860,270 priority patent/US5959663A/en
Publication of WO1997015150A1 publication Critical patent/WO1997015150A1/ja

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T15/003D [Three Dimensional] image rendering
    • G06T15/10Geometric effects
    • 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
    • H04N13/111Transformation of image signals corresponding to virtual viewpoints, e.g. spatial image interpolation
    • H04N13/117Transformation of image signals corresponding to virtual viewpoints, e.g. spatial image interpolation the virtual viewpoint locations being selected by the viewers or determined by viewer tracking
    • 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
    • H04N13/167Synchronising or controlling 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
    • H04N13/10Processing, recording or transmission of stereoscopic or multi-view image signals
    • H04N13/194Transmission of 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
    • H04N13/261Image signal generators with monoscopic-to-stereoscopic image conversion
    • 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
    • H04N13/30Image reproducers
    • H04N13/398Synchronisation thereof; Control thereof
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N5/00Details of television systems
    • H04N5/74Projection arrangements for image reproduction, e.g. using eidophor
    • 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
    • H04N13/30Image 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]
    • H04N13/344Displays for viewing with the aid of special glasses or head-mounted displays [HMD] with head-mounted left-right displays

Definitions

  • the present invention relates to a method and an apparatus for generating a three-dimensional image, and more particularly to a method and apparatus for generating a three-dimensional image that allows an operator to perceive a three-dimensional image in a virtual space based on a video signal representing a two-dimensional image. is there. Background art
  • a stereoscopic video display device as shown in FIG. 1 has been proposed as a stereoscopic video generation device that allows an operator to view a stereoscopic video.
  • the three-dimensional image display device 1 uses the left-eye projector 2L and the right-eye projector 2R to convert the left-eye projection light and the right-eye projection light into red, green, and blue primary color polarization filters 3LR, 3LR, respectively.
  • a color composite image is projected on the screen 4, and the color composite image is displayed for the left and right eyes. It is configured so that an operator can view it through the glasses 5 having 5 L and 5 R for use.
  • the operator observes the left-eye image projected from the left-eye projector 2L through the left-eye filter 5L of the color composite image and simultaneously projects the left-eye image from the right-eye projector 2R through the right-eye filter 5R.
  • the right-eye image is viewed, and as a result, the color composite image on the screen 4 is displayed based on the parallax between the left eye 6L and the right eye 6R as shown in FIGS. 2 (A) to 2 (C). And perceive it as a stereoscopic image.
  • the line of sight 8L of the left eye 6L and the line of sight 8R of the right eye 6R intersect at the same position on the screen 4. It is perceived as “there is a composite stereoscopic image on screen 4”.
  • the left-eye image 7L and the right-eye image 7R are shifted from each other at a position facing the left eye 6L and the right eye 6R on the screen 4.
  • a point in the virtual space where the left eye 6L intersects with the line of sight 8L looking at the left-eye image 7L and the right eye 6R intersects with the line of sight 8R looking at the right-eye image 7R (
  • the virtual image 9B is formed on the screen 4 (which occurs behind the screen 4)
  • the operator perceives that "the composite stereoscopic image is located behind the screen 4".
  • the operator can perform the screen operation.
  • the composite stereoscopic image can be viewed at a position on the screen 4 or at a position behind or in front of the screen 4.
  • the parallax between the left eye 6L and right eye 6R is not so large (about 3.25 [cm]).
  • Pellet One has to perform a complicated setting operation of setting the projection positions of the left-eye image 7L and the right-eye image 7R while changing the projection positions by trial and error. Disclosure of the invention
  • the present invention has been made in view of the above points, and can easily project a composite stereoscopic image of a left-eye image and a right-eye image at a desired position on a screen as needed. It is intended to propose a stereoscopic video generation method and apparatus that can be used.
  • a source video signal is subjected to image conversion processing by a left-eye image conversion processing unit and a right-eye image conversion processing unit, so as to cope with parallax on a screen.
  • a left-eye image and a right-eye image can be generated at desired positions.
  • a spatial image conversion matrix for converting an image represented by the source video signal to a virtual spatial position and a converted video signal converted to the virtual spatial position are used.
  • An image conversion operation for generating a left-eye stereoscopic video signal from a source video signal is performed based on a left-eye perspective conversion matrix that allows the converted video to be viewed on a screen surface.
  • a source video signal is obtained based on the spatial image conversion matrix and a right-eye perspective conversion matrix for seeing the converted image represented by the converted video signal on a screen surface.
  • An image conversion operation is performed to generate a stereoscopic video signal for the right eye from.
  • the left-eye image and the right-eye image can be made to have an appropriate parallax depending on the parameters input by the operator. This makes it possible to easily generate a stereoscopic video with higher image quality.
  • a desired three-dimensional transformation of an operating system can be performed on a two-dimensional source video signal.
  • the source video Even if the signal is moved continuously for each frame, the right-eye video signal and the left-eye video signal can be generated in real time in conjunction with the continuous movement of the source video signal. .
  • the image conversion processing is performed simultaneously based on one input video signal, which makes it difficult to resynchronize the left-eye and right-eye images. There is no need to generate a higher quality stereoscopic video. Therefore, the left-eye video signal and the right-eye video signal that are always synchronized can be displayed on the screen.
  • FIG. 1 is a perspective view showing a stereoscopic video display device to which the present invention can be applied.
  • 2 (A) to 2 (C) are schematic diagrams for explaining that an operator can perceive a stereoscopic image by the configuration of FIG.
  • FIG. 3 is a schematic diagram for explaining a left-eye image and a right-eye image to be generated by the present invention.
  • FIGS. 4 (A 1) to 4 (B 2) are schematic diagrams for explaining the operation principle of the stereoscopic video image generating apparatus and method according to the present invention.
  • FIG. 5 is a schematic diagram for explaining the image conversion processing between the position vector on the frame memory and the position vector on the screen.
  • FIG. 6 is a block diagram showing a first embodiment of the stereoscopic video image generating apparatus according to the present invention.
  • FIG. 7 is a block diagram showing a second embodiment of the stereoscopic video generation device according to the present invention.
  • FIGS. 8A and 8B are schematic diagrams used to explain the operation of FIG. BEST MODE FOR CARRYING OUT THE INVENTION
  • the method of generating a stereoscopic image according to the present invention is as follows.
  • the object 13 which is a synthetic stereoscopic image having a depth at an arbitrary position
  • the image of the object 13 is displayed on the left eye.
  • the images 14L and 14R for the left and right eyes seen through the screen 11 along the line of sight of the 12L and the right eye 12R are obtained by an image conversion operation.
  • the depth means information representing a position in a direction along an intermediate line passing through a center position between the lines of sight of the left eye 12L and the right eye 12R looking at the object 13.
  • a video V 1 (referred to as an object) represented by an input video signal as a source video signal is a screen on the X and y axes. It is set as the image projected on the origin position of the Xy plane on the screen surface of 11.
  • This object V1 which is the video of the input video signal input as the source video signal, is expressed by the following equation:
  • the conversion process using the three-dimensional image conversion matrix ⁇ ⁇ is performed by using the spatial image conversion matrix ⁇ ⁇ ⁇ ⁇ for the input video signal.
  • the image of the object VI represented by the input video signal is translated in the depth direction along the z-axis from the position on the x, y plane on the screen 11 After rotation, it means that the converted image has depth information at the position behind the screen 11.
  • This intermediate conversion image V 2 is used for image conversion representing parallax from the left eye 12 L.
  • the conversion process is performed by the + ⁇ movement matrix L represented by ⁇ 01, whereby an intermediate converted video V 3 L is obtained.
  • This + ⁇ movement matrix L means that a conversion operation for moving the intermediate converted image V 2 by a distance + ⁇ X in parallel with the direction along the X-axis is performed, and thus the left eye 1 2 This means that the intermediate conversion video V 3 L has been viewed from the position 1 2 LX of the left eye 1 2 L, assuming that it is on the string.
  • the perspective matrix ⁇ represented by 0 0 1.
  • An intermediate conversion image V 4 L is obtained on the screen 11 on the X and y axes.
  • This one ⁇ X movement matrix L— 1 translates the intermediate converted image V 4 L in parallel in the X-axis direction by a distance of one ⁇ X, and thus the final converted image V 5 L
  • the intermediate converted image V 2 By moving the position of the left eye 12 L to the left from the center position by ⁇ as the intermediate position between the left eye 12 L and the right eye 12 R, the intermediate converted image V 2 from the original position of the left eye 12 L When viewed, it is obtained as an image seen through the screen 11 (Fig. 4 ( ⁇ 1)).
  • the left-eye image 14L represents a perspective image in which the intermediate converted image V2 has a depth, so that the image part in the foreground looks large and the image part behind in the depth direction looks small. become.
  • conversion parameter Isseki r used in formulas, and ⁇ r 3 3 a parameter Isseki for rotating the input image V 1 x, in three-dimensional space of y and z coordinates, also transformation parameter, I y and £ z is, x, a parameter Isseki for moving parallel to the y and z directions, transformation parameters S t, is a parameter for magnifying or reducing the input image V 1 2 dimensionally is there.
  • the conversion parameters used in equations (2) and (4), ⁇ X and I ⁇ X are parameters for translating the intermediate converted image in the X direction by a distance of ten ⁇ X and one ⁇ X. It is an overnight.
  • the value of 16 is set to be a reference value. (This means that the operator's viewpoint is at position –16 of the z coordinate.
  • the first conversion processing for converting the object V1 into the intermediate converted video V2 before obtaining the final converted video V5L for the left eye based on the object V1 which is the video of the input video signal Focusing on the steps and a second conversion processing step of performing a conversion process for obtaining the left-eye video 14 L based on the intermediate converted video V 2, the left-eye conversion matrix TL is
  • T T.
  • the transformation matrix P L used in the second transformation processing step is
  • Equation (6) the left-eye transformation matrix T
  • L is r r 1 1 r i r 1 3 0 1 0 0 0) r 2 1 r 2 r 2 3 0 0 1 0 0
  • the transformation matrix TL expressed by the equation (9) is actually the input video signal After the video data has been loaded into the frame memory, the conversion process is executed according to the parameters constituting the matrix, but the data read from the frame memory and displayed on the screen 11 Since the video data to be obtained is two-dimensional data, the parameters in the third row and the third column of the equation (9) are not necessary when calculating a two-dimensional address.
  • Transformation matrix T L33 Chikaraku represented by the position vector preparative file on a frame memory [x, y] position base-vector on disk rie down 1 1 [X S, Y s] for conversion to It will be used as a matrix.
  • the transformation matrix of 4 rows and 4 columns as shown in the equations (1) to (4) expresses different transformations of rotation transformation and enlargement and reduction in the same coordinate system.
  • This is a matrix of 4, which is generally called a homogeneous coordinate system (Homogeneous Coordinate).
  • two-dimensional position base-vector [X, y] on full Remume Mori and disk rie down 1 1 on the two-dimensional position base-vector [X S, Y s] homogeneous coordinates expressed in the system can be expressed as [x, y] is [x, can be expressed y, H] and, [X s, Y s] is [X s, Y s, 1].
  • H is a value representing the image magnification or reduction ratio transformed by the perspective method, and is proportional to the value of the position vector in the z-axis direction.
  • the input video data is read out from the frame memory so that the input video data is three-dimensionally converted.
  • the address specify the address on the frame memory.
  • each element of the transformation matrix T L33 is represented by r 11-r 13 P um xr, 3 ⁇ ⁇
  • the position vector [x, y, H] on the frame memory is a vector of the homogeneous coordinate system, to return it to the two-dimensional coordinate system, the coordinate H What is necessary is just to normalize the values x and y.
  • H ⁇ r (L % -£, P ⁇ ⁇ ⁇ ) + (r 2 r 23 P, ⁇ x) SL y ⁇ X + ir 1 2 (2, — ⁇ ⁇ ⁇ ⁇ x x)-(r , r, 3 P, ⁇ x) ⁇ y ⁇ Y + ⁇ -r (r-r P, ⁇ x) + (r-r P! A x) r
  • the object V 1 input as the image of the input video signal at the origin of the xy plane on the screen 11 is represented on the Z axis by the The image is translated in the direction and rotated to be converted into the intermediate converted image V 2.
  • the intermediate transformed video V 3 R is transformed by the perspective matrix P expressed by the equation (3), whereby the intermediate transformed video V 3 R is seen through the screen 11.
  • Video V 4 R is obtained.
  • the intermediate conversion image V 4 R is seen through the intermediate conversion image V 2 on the screen 11 in the line of sight of the right eye 12 R. It means that you got the video V4R.
  • This intermediate conversion video V 4 R is represented by the following equation (2).
  • the image is translated in the X-axis direction by a distance + ⁇ X and converted into the final converted video V5R for the right eye.
  • the final converted video V5R for the right eye has the right eye 12R shifted from the center position on the z-axis (the center position of the left eye 12L and the right eye 12R) by ⁇ X to the right.
  • the intermediate converted image V 2 is viewed by the right eye 12 R at the original position, the image obtained by seeing through the intermediate converted image V 2 on the screen 11 in the original line-of-sight direction of the right eye 12 R It represents that.
  • the final converted video V5R for the right eye has a depth as shown in Fig. 4 (B2).
  • a right-eye image 14 R on the xy plane that has been processed based on perspective so that the front image part is large and the depth image part is small. You can do it.
  • the conversion process of converting the object V 1 based on the input video signal into the final converted video V 5 R for the right eye is performed as follows, similar to the conversion process for the left eye .
  • equation (51) to obtain significant information only in the xy plane, the following equation excluding the parameters in the third row and third column of equation (51) is used: r!, + r, P 2 ⁇ xrr l 3 P z
  • [X Y 1] C X y H] T (53) is defined by ⁇ :.
  • the position base-vector on disk rie down 1 1 [X S Y s 1] is designated in accordance with the raster scan
  • the position vector [xy H] of the corresponding frame memory is specified, thereby obtaining spatial image transformed two-dimensional video data and a key signal.
  • the position vector [xy H] of the frame memory can be reduced.
  • H b 31 Xs + b Y s + b ⁇ can be determined by the (84).
  • H ⁇ -r 2 (J2 x + S. z P z A x) + (r 21 + r 23 P z A x)-£ y ⁇ X + ⁇ r, 2 (Jg x + S, z P 2 A x)-(r,, + r, 3 P, ⁇ x)- ⁇ y ⁇ Y + (-1-1 2-(r 2 1 + r 23 P 2 ⁇ x) + (r,, + r, 3 P, ⁇ x) r 2 2 i
  • the intermediate VI obtained by moving and rotating the object VI represented by the input video signal in the depth direction and in the ⁇ and ⁇ directions.
  • the converted images V 3 L and V 3 R By seeing the converted images V 3 L and V 3 R on the screen 11 in a direction along the line of sight of the left eye 12 L and the right eye 12 R, respectively, the left eye 12 L and the right eye 12 R
  • the final converted images V5L and V5R to be projected on the screen 11 with the parallax between them can be obtained, whereby the depth can be increased based on the parallax of the left eye 12L and the right 12R.
  • An image of the object can be generated on the screen surface, at a position behind the screen surface, or at a position in front of the screen surface.
  • the generation of the stereoscopic image is represented by a three-dimensional transformation matrix ⁇ 0 defined by the equation (1), a + ⁇ X movement matrix L defined by the equation (2), and a perspective matrix defined by the equation (3).
  • P. And (4) is defined by the formula - delta by each path lame Isseki operator inputs defining the X movement matrix L 1, facilitates the three-dimensional image having an arbitrary conversion corresponding to an arbitrary position as required Can be generated.
  • the table Diary to write, as a parameter that can be set, the table to parameters Ichita r a spatial rotation of converted image,, ⁇ r 33 and, X-axis, parameter one represents the translation of the y-axis and z-axis direction evening J2 ,, S, y and J2, parameters representing two-dimensional expansion / compression, parameters Pz representing perspective values based on perspective, and half the distance between the left eye 12 L and the right eye 12 R Since the value ⁇ X indicating the distance can be arbitrarily selected and designated as needed, the video to be generated can be easily specified.
  • the read address can be specified in correspondence with the raster scan address on the screen, so that the screen 11
  • the final converted images V5L and V5R for the left and right eyes upward can be simultaneously generated.
  • reference numeral 21 denotes a stereoscopic image generating apparatus according to the present invention as a whole.
  • the input video signal V A , the input key signal K A, and the input background signal V BK are transmitted to the left eye processor 22.
  • the left-eye and right-eye projectors 23L and 23R project the left-eye and right-eye final converted images V5L and V5R, respectively, on the screen 24.
  • the left-eye processor 22 L and the right-eye processor 22 R, and the left-eye and right-eye projectors 23 L and 23 R are controlled by a central processing unit (CPU) 32 via a bus 31.
  • CPU central processing unit
  • the CPU 32 executes the processing operation of each part of the three-dimensional image generation device 21 according to the program in the program memory 33 having the ROM configuration and using the single memory 34 having the RAM configuration.
  • the CPU 32 generates write address signals SX and Sy representing the position vector [xy] on the frame memory from the write address generation circuit 35, and the write address signals Sx and Sy.
  • the input video signal VA is used for the left eye and the right eye. Synchronize with each other as source video data and store them.
  • the disturbing address signals S x and S y are also supplied to the key signal memories 37 L and 37 R of the left-eye and right-eye processors 22 L and 22 R, thereby synchronizing with the input video signal V A.
  • the key data for the position vector [xy] on the frame memory is stored in the key signal memories 37L and 37R.
  • the CPU 32 drives the screen address generation circuit 38 provided in common for the left-eye and right-eye processors 22 L and 22 R, thereby causing the position vector (X S Y s] is generated based on the reference signal generated from the built-in reference signal generation circuit, and the screen address signals SX s and SY s are generated by the left-eye and right-eye processors 22 L and 22 R. It is supplied to the read address generator 39 L and 39 R.
  • the read address generation circuits 39 L and 39 R respectively perform the read operations specified by the equations (45), (46), (87), and (88).
  • Read address signals SX Lm , SY and SX Rm , SY Rm representing address (Xt m Y Lm ) and (X Rm Y Rm ) are stored in frame memories 36 L and 36 R, respectively. It is supplied to 37 L and 37 R.
  • the video data VL and VR read from the frame memories 36 L and 36 R of the left-eye and right-eye processors 22 L and 22 R are video signal interpolation circuits 40 L and 4 L, respectively. After being subjected to interpolation processing at 0 R, they are given to the combiners 41 L and 41 R as read video signals VL and VR.
  • the key data KL and KR, read from the key signal memories 37 and 37R of the left-eye and right-eye processors 22L and 22R, respectively, are used as the key signal interpolation circuit 42L and after being interpolated at 4 2 R, provided to co Npaina 4 1 L and 4 1 R as a read key signal KL 2 and KR.
  • Combiner 4 1 L and 4 1 R under the control of the CPU 3 2, by reading the key signal KL 2 and KR 2, the read video signal VL 2 and VR 2, the input back Gras ⁇ down de signal V BK The keying process between VLKL 2 VL 2 + (1-KL) V BK (90)
  • VR U KR 2 VR 2 + (1-KR 2) V BK A stereoscopic video output signal VL represented by (91).
  • ut and VR. ut is supplied to the left-eye and right-eye projectors 23 L and 23 R as outputs of the left-eye and right-eye processors 22 L and 22 R.
  • the operator inputs the values of the parallax distance ⁇ X and the perspective value Pz required in advance for the calculation of the present invention via the control panel 51.
  • the CPU 32 stores the parallax distance ⁇ X and the perspective value P, which are newly set by the operator, in the RAM 34.
  • the operator instructs a three-dimensional image conversion operation on the source video signal by operating a three-dimensional pointing device or the like provided on the control panel 51.
  • the CPU 32 When the operator instructs the source video signal to perform some kind of three-dimensional conversion, the CPU 32 generates a determinant ⁇ representing the three-dimensional conversion specified by the operator. , £ t and S N , and at the same time, the parallax distance ⁇ X and the perspective value P: stored in the RAM 34.
  • the CPU 32 receives the received parameters i-,, ⁇ r 33 ⁇ , i y i t S. Using ⁇ ⁇ and, the values of the parameters b, b 33 shown in equations (35) to (43), and the parameters b,!
  • the read address generation circuit 39 R is connected to the parameters bll to b 33 supplied from the CPU 32 and the screen address supplied from the screen address generation circuit 38 ( X Lm , Y Lm ), and generates a right-eye read address (X Rm , Y Rm ) as shown in Expression (85).
  • Read ⁇ address generating circuit 3 9 R is generated read address for the right eye (X Rm, Y Rm) to be Kyoawase the full Remume mode Li 3 6 R and the key signal memory 3 7 R.
  • the converted video signal VL, read from the frame memory 36L has the spatial conversion matrix T as shown in FIG. 4 (A1). Shows the video signal V5L projected onto the screen when the video signal V2 converted into a three-dimensional space is viewed from the viewpoint position 12L of the left eye.
  • the converted video signal VR, read from 36 R is converted from the video signal V 2, which is converted into a three-dimensional space by the spatial The video signal V5R projected onto the screen when viewed from the position 12R is shown.
  • the converted video signals VL and VR read from the frame memories 36 L and 36 R are interpolated by peripheral pixels by the video signal interpolators 40 L and 40 R, and converted video signals are output.
  • the converted key signals KL and KR read from the key signal memories 37 L and 37 R, are interpolated by peripheral pixels by the key signal interpolators 42 L and 42 R. is processed and outputted as a converted key signal KL 2 and KR 2.
  • the CPU 32 captures the input information input by the operator from the control panel 51 through the interface 52, thereby performing the image conversion processing specified by the operator.
  • Input video signal VA and input key signal K A stereoscopic video output signal VL for the left and right eyes, which is obtained by simultaneously converting A. ut and VR ou, are supplied from the left-eye and right-eye processors 22 L and 22 R to the left-eye and right-eye projectors 23 L and 23 R.
  • Isseki parameters from an operator co emission control panel 5 1 1, if set the ⁇ r 33, CPU 3 2 is 3-dimensional transform in 9 gun out ⁇ address generating circuit 3 L and 3 9 R Matrix P. For, specify how spatially the object V 1 should be rotationally transformed.
  • the CPU 32 generates the three-dimensional conversion matrix P in the read address generation circuits 39L and 39R. For, specify how much the object VI should be translated in the x $, y, and z directions.
  • the operator presses the control panel 51 from the control panel 51.
  • the CPU 32 generates a three-dimensional transformation: transformation matrix P in the read address generation circuits 39 L and 39 R. Specifies the extent to which object V 1 should be scaled up or down on the xy plane.
  • the CPU 32 causes the read address generation circuits 39L and 39R to add + ⁇ movement matrix L and one ⁇ ⁇ .
  • the distance X ie the disparity between the left eye 12 L and the right eye 12 R
  • SX Lm , SY Lm, SX Rm , and SY Rm are generated.
  • a value of 1 to 16 is always input from the control panel 51, and thus the CPU 32 reads the parameter.
  • the dress generation circuits 39 L and 39 R a read address signal for defining a perspective value based on the perspective method to a constant value is transmitted for the perspective matrix P n .
  • the operator can operate the control panel 51 from the left eye 12 L and the right eye 12 L Since the parameter ⁇ x that defines the parallax of R can be selected to a desired value, the final converted image V 5 L projected onto the screen 24 by the left-eye and right-eye projectors 23 L and 23 R and The amount of deviation between the left-eye image 14L and the right-eye image 14R of V5R can be easily set to an appropriate value according to preference.
  • the processing of the read address generation circuits 39 L and 39 R of the left-eye and right-eye processors 22 L and 22 R is performed simultaneously based on the common screen address signals SX s and SY S.
  • the stereoscopic video output signal VL for the left eye and the right eye because it is executed.
  • ut and VR. The ut is simultaneously generated in time, so that a stereoscopic image with higher image quality can be generated.
  • the source video signal VI is converted to a spatial position by moving and rotating the source video signal VI in the three-dimensional space by using the left-eye perspective transformation matrix PL and the right-eye perspective transformation matrix in the calculation.
  • a left-eye video signal and a right-eye video signal are calculated by projecting the converted video signal onto the xy plane from the left eye position 12 L and the right eye position 12 shaku. Therefore, it is possible to generate an appropriate video signal according to the parallax between the left eye and the right eye.
  • the more the source video signal is three-dimensionally converted to a position in the depth direction (+ z direction) with respect to the screen the more the left-eye video signal shifts to the left and the right-eye video signal shifts. It is calculated so that it shifts to the right.
  • the more the source video signal is converted to a position closer to the screen ( ⁇ z direction) the more the left-eye video signal shifts to the right and the more the right-eye video signal shifts. It is calculated so as to shift to the left.
  • a video signal having a three-dimensional effect can be easily generated.
  • FIG. 7 shows a second embodiment of the present invention.
  • the stereoscopic image generating apparatus 61 assigns the same reference numerals to the parts corresponding to those in FIG.
  • Left-eye and right-eye stereoscopic video output signals VL obtained from the clusters 62L and 62R. UT and VR. ut is output to the left-eye and right-eye projectors 23 L and 23 R, respectively, and is recorded on the left-eye and right-eye video signal recording devices 63 L and 63 R, which are, for example, video tape recorders. I have.
  • the left-eye and right-eye video signal recorders 633 and 6713 ⁇ 4 reproduce the recorded video signals and switch them as left-eye and right-eye feedback video signals VBL and VBR, respectively.
  • the video input signal VL to the frame memories 36 L and 36 R of the left eye and right eye processors 62 L and 62 R through the switching inputs W 2 of the switches 64 L and 64 R. And VR. And return.
  • the input video signal VA is passed through the first switching input terminal W 1 of the left-eye and right-eye switching switches 64 L and 64 1 ⁇ that perform switching operation under the control of the CPU 32.
  • WR. Is input to the frame memories 36 L and 36 R.
  • the input video signals VL0 and VR input through the left-eye and right-eye switching switches 64L and 64R.
  • VR n synchronized with the input key signal may be a keying process KL.
  • KRQ are input to the key signal memories 37 L and 37 R. In the configuration shown in FIG. 7, as shown in FIGS.
  • the first left-eye and right-eye images 70 L and 7 OR are projected on a part of the screen 24 (for example, the left part) and the other part of the screen 24 (for example, the right part).
  • the second left-eye and right-eye images 72 L and 72 R are projected.
  • the first left-eye and right-eye images 70L and 70R are images represented by the input video signal VA , and the method described above with reference to FIGS. 4 (A1) to 4 (B2). Perform image conversion processing with.
  • the first left-eye and right-eye images 70L and 70R form a virtual image in which the eyes of the left eye 12L and the right eye 12R are located behind the screen 24.
  • the stereoscopic image 71 formed based on the input video signal VA is perceived by the operator.
  • the second left-eye and right-eye images 72 L and 72 R projected on the right portion of the screen 24 are output from the left-eye and right-eye video signal recording devices 63 L and 63 R.
  • the reproduced video signals represented by the left-eye and right-eye feedback video signals V 8 and 81 ° are subjected to image conversion processing by the image conversion method described above with reference to FIGS. 4 (A 1) to 4 (B 2).
  • the read address signals SX Lm , SY Lm, SX Rm , and SY Rm are read from the video memories 36 L and 36 R and projected onto the screen 24.
  • Fig. 8 (A) the lines of sight of the left eye 12L and the right eye 12R with respect to the second left-eye and right-eye images 72L and 72R intersect before the screen 24.
  • the stereoscopic image 73 formed by the left-eye and right-eye return video signals VBL and VBR in the virtual space in front of the screen 24 is perceived by the operator.
  • the left-eye and right-eye switching switches 64L and 64R are input under the control of the CPU 32 at the timing of the left part of the video section for one frame.
  • the video signal VA is stored in the left memory area of the left-eye and right-eye frame memories 36L and 36R through the first switching input terminal W1.
  • the left-eye and right-eye switching switches 64 L and 64 R are switched to the second switching input terminal W 2 during the left portion of one frame period, so that the left-eye and right-eye switching switches are switched.
  • the feedback video signals VBL and VBR are stored in the frame memories 36L and 36R.
  • one frame of video data including two video data is stored in the frame memories 36 L and 36 R, and the video data is stored under the control of the CPU 32.
  • the output address generation circuits 39L and 39R generate the readout address signals that have been subjected to the image conversion processing, thereby generating the first left-eye and right-eye video signals.
  • No. 7 OL and 7 OR and the left-eye and right-eye stereoscopic video output signals VL each consisting of one frame of video representing the second left-eye and right-eye images 72 L and 72 R. ut and VR. It sends the U t in the projector 2 3 L and 2 3 R.
  • UL and VR. ut is stored in the left-eye and right-eye storage devices 63L and 63R, and only the video signals of the second left-eye and right-eye images 72L and 72R are switched between the left-eye and right-eye switching switches.
  • VR This is stored in the frame memories 36L and 36R.
  • the projectors 23L and 23R perform the repeated image conversion processing on the second left-eye and right-eye images 72L and 72R projected on the right portion of the screen 24.
  • the stereoscopic video 73 formed by the feedback video signals VBL and VBR changes sequentially for each frame.
  • the three-dimensional image 71 formed by the first left-eye and right-eye images 70 L and 70 R based on the input video signal VA whose image content does not change does not change. Perceived by the operator as a fixed image.
  • the perspective matrix P In the above embodiment, the perspective matrix P. In the above, the case where the parameter Pz in Equation (3) is fixed to 1/16 in advance to keep the perspective value based on the perspective method unchanged, but the present invention is not limited to this. The same effect as in the above-described embodiment can be obtained even when the control panel 51 can be changed by the operator's operation on the control panel 51.
  • a part of the image projected on the screen 24 is returned to the frame memories 36 L and 36 R.
  • the present invention is not limited to this, and the image conversion process may be repeatedly performed for all the images generated on the screen 24.
  • the left-eye and right-eye stereoscopic video output signals VL. ut and VR. ut is projected on the screen 24 by the left-eye and right-eye projectors 23L and 23R to generate a stereoscopic image, but as a means for displaying the stereoscopic image.
  • the present invention is not limited to this, and another type of video display device such as a head mount type video device may be applied.
  • the video tape recorder is applied as the left-eye and right-eye video signal recording devices.
  • the video signal recording device is not limited thereto.
  • Another video recording device such as a video disk device or a semiconductor memory device may be used.
  • the present invention is applicable to a stereoscopic video generation device that generates a stereoscopic video using a video signal representing a two-dimensional video.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • Geometry (AREA)
  • Computer Graphics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)
  • Studio Circuits (AREA)

Description

明 細 書 発明の名称
立体映像生成方法及び装置 技術分野
本発明は立体映像生成方法及び装置に関し、 特に 2次元的な映像を表わすビデ ォ信号に基づいて、 オペレータが仮想空間内に立体的な映像を知覚できるような 立体映像を生成しょうとするものである。 背景技術
従来、 オペレータが立体映像を観賞できるようにした立体映像生成装置と して 、 図 1 に示すような、 立体映像表示装置が提案されている。
.この立体映像表示装置 1 は、 左目用プロジェク タ 2 L及び右目用プロジェク タ 2 Rにおいて、 左目用投影光及び右目用投影光を、 それぞれ赤、 緑、 青 3原色偏 向フィルタ 3 L R、 3 L G、 3 L B及び 3 R R、 3 R G、 3 R Bを通して、 スク リー ン 4 に投射することにより、 当該スク リ ーン 4上に力ラー合成映像を映出し 、 当該カラー合成映像を左目用及び右目用フィル夕 5 L及び 5 Rをもつ眼鏡 5を 通してオペレータが目視するように構成されている。
この場合オペレータは、 カラー合成映像のうち、 左目用フィルタ 5 Lを通して 左目用プロジェクタ 2 Lから投射された左目用映像を目視すると同時に、 右目用 フィル夕 5 Rを通して右目用プロジェクタ 2 Rから投射された右目用映像を目視 し、 その結果、 スク リ ーン 4上のカラー合成映像を、 図 2 ( A ) 〜図 2 ( C ) に 示すように、 左目 6 L及び右目 6 R間の視差に基づいて、 立体映像と して知覚す る。
因に、 図 2 ( A ) に示すように、 スク リ ーン 4上の同一位置に左目用プロジェ ク タ 2 Lから投射された左目用映像 7 Lと、 右目用プロ ジェク タ 2 Rから投射さ れた右目用映像 7 Rとが映出された場合、 左目 6 Lの視線 8 Lと、 右目 6 Rの視 線 8 Rとがスク リ ーン 4上の同一位置において交差するので、 オペレータは 『ス ク リ ーン 4上に合成立体映像がある』 ように知覚する。
これに対して、 図 2 ( B ) に示すように、 スク リ ーン 4上の左目 6 L及び右目 6 Rと対向する位置に、 左目用映像 7 L及び右目用映像 7 Rが互いにずれた状態 で映出されたときには、 左目 6 Lが左目用映像 7 Lを目視する視線 8 Lと、 右目 6 Rが右目用映像 7 Rを目視する視線 8 Rとが交差する仮想空間内の点 (スク リ —ン 4 より後方に生ずる) に虚像 9 Bが結像されることにより、 オペレータは 『 スク リ ー ン 4 の後方位置に合成立体映像がある』 ように知覚する。
また、 図 2 ( C ) に示すように、 スク リーン 4上の左目 6 L及び右目 6 Rと対 向する位置に、 右目用映像 7 R及び左目用映像 7 Lを投射すると、 左目 6 Lが左 目用映像 7 Lを目視する視線 8 Lと右目 6 Rが右目用映像 7 Rを目視する視線 8 Rとが仮想空間内のスク リーン 4 より前方位置において交差することにより虚像 9. Fが結像されるので、 オペレータは 『スク リ ーンの前方位置に合成立体映像が ある』 ように知覚する。
従って、 左目用映像 7 L及び右目用映像 7 Rのスク リ ーン 4上の投射位置を、 左目 6 L及び右目 6 Rの視差に相当する距離だけずらせるようにすることにより 、 オペレータはスク リ ーン 4上の位置、 又はスク リ ーン 4の後方又は前方位置に 合成立体映像が映出されているものと して観赏することができる。
と ころがこのような手法によって立体映像を生成しょう とする場合、 左目用プ ロジェクタ 2 L及び右目用プロジェクタ 2 Rから投射されている左目用映像 7 L 及び右目用映像 7 Rの投射位置を決めることが、 実際上かなり難し く 、 オペレー 夕にかなりの熟練度が要求される。
因に、 実際上左目 6 L及び右目 6 Rの視差はそれほど大き く はないので (3. 25 [ cm) 程度である) 、 実際に効果的な立体映像を生成するためには、 熟練したォ ペレ一夕が左目用映像 7 L及び右目用映像 7 Rの投射位置を、 試行錯誤的に変更 しながら設定する、 という煩雑な設定作業をしなければならない問題がある。 発明の開示
本発明は以上の点を考慮してなされたもので、 左目用映像及び右目用映像の合 成立体画像を、 必要に応じてスク リ ーン上の所望の位置に容易に映出させること ができるようにした立体映像生成方法及び装置を提案しょう とするものである。 かかる課題を解決するため本発明においては、 ソースビデオ信号を、 左目用画 像変換処理手段及び右目用画像変換処理手段によって画像変換処理をすることに より、 スク リ ーン上の視差に対応する所望の位置に、 左目用映像及び右目用映像 を生成できるようにする。
左目用映像を得るために、 ソースビデオ信号によって表される映像を仮想的な 空間位置に変換するための空間的画像変換行列と、 当該仮想的な空間位置に変換 された変換ビデオ信号によって表される変換映像をスク リ ーン面上に透視する左 目用透視変換行列とに基づいて、 ソースビデオ信号から左目用立体ビデオ信号を 生.成するような画像変換演算が実行される。 また右目用映像を得るために、 上記 空間的画像変換行列と、 上記変換ビデオ信号によって表される変換映像をスク リ ーン面上に透視する右目用透視変換行列とに基づいて、 ソースビデオ信号から右 目用立体ビデオ信号を生成するような画像変換演算が実行される。
かく して本発明によれば、 2次元的な映像を表す入力ビデオ信号に基づいて、 左目用及び右目用映像を、 オペレータが入力したパラメ一夕によって適切な視差 をもつようにすることができることにより、 一段と画質が良好な立体映像を容易 に生成することができる。
また、 本発明によれば、 空間的変換行列と左目用透視変換行列及び右目用透視 変換行列を使用することによって、 2次元的なソースビデオ信号に対して、 オペ レー夕の所望の 3次元変換処理を施したときに、 変換されたソースビデオ信号の 空間的な位置に基づいて、 適切な視差を有した左目用の 2次元的なビデオ信号と 右目用の 2次元的なビデオ信号を、 演算によって容易に生成することができる。 さらに、 3次元空間内において、 オペレータの操作に基づいて、 ソースビデオ 信号をフ レーム毎に連続的に移動させたと しても、 そのソースビデオ信号の連続 的な動きに連動して、 リアルタイムで右目用のビデオ信号及び左目用のビデオ信 号を生成することができる。
また左目用及び右目用映像を生成する際に、 1 つの入力ビデオ信号に基づいて 同時に画像変換処理を実行するようにしたことにより、 左目用及び右目用映像の 同期を取り直すといった面倒な処理をする必要がなく 、 この分一段と良質な立体 映像を生成することができる。 よって、 スク リーン上に、 常に同期の取れた左目 用のビデオ信号と右目用のビデオ信号を表示することができる。 図面の簡単な説明
図 1 は本発明を適用し得る立体映像表示装置を示す斜視図である。
図 2 ( A ) 〜図 2 ( C ) は図 1 の構成によってオペレータが立体映像を知覚で きることを説明するための略線図である。
.図 3 は本発明によって生成しよう とする左目用及び右目用映像の説明に供する 略線図である。
図 4 ( A 1 ) 〜図 4 ( B 2 ) は、 本発明による立体映像生成装置及び方法の動 作原理を説明するための略線図である。
図 5 はフ レームメモリ上の位置べク トルとスク リ ーン上の位置べク トル間との 画像変換処理の説明に供する略線図である。
図 6 は本発明による立体映像生成装置の第 1実施例を示すブロック図である。 図 7 は本発明による立体映像生成装置の第 2実施例を示すブロック図である。 図 8 ( A ) 及び図 8 ( B ) は図 7の動作の説明に供する略線図である。 発明を実施するための最良の形態
以下図面について、 本発明の一実施例を詳述する。
( 1 ) 立体映像の生成原理
本発明による立体映像の生成の仕方は、 図 3 に示すように、 スク リ ー ン 1 1 を 含む x y z仮想空間において、 左目 1 2 L及び右目 1 2 Rから見て任意の位置に 、 奥行きをもった合成立体映像である物体 1 3を見たときに、 当該物体 1 3の映 像を左目 1 2 L及び右目 1 2 Rの視線に沿ってスク リ ーン 1 1 上に透視した左目 用及び右目用映像 1 4 L及び 1 4 Rを画像変換演算によって求める。
ここで奥行きとは、 物体 1 3を見る左目 1 2 L及び右目 1 2 Rの視線間の中央 位置を通る中間線に沿った方向の位置を表わす情報を意味する。
図 4 ( A 1 ) に示すように、 ソースビデオ信号と しての入力ビデオ信号によつ て表される映像 V 1 (これをオブジェク トと呼ぶ) は、 X及び y軸上にあるスク リーン 1 1 のスク リ一ン面上の X y平面の原点位置に映出された映像と して設定 される。
ソースビデオ信号と して入力された入力ビデオ信号の映像であるこのォブジェ ク ト V 1 は、 次式
「 Γ r r 0
1" I Γ 0
T ( 1 )
r Γ Γ 0
、 £ x s によって表される 3次元画像変換行列 T。 によって平行移動演算処理されると共 に回転演算処理されて中間変換映像 V 2 に変換される。
かかる 3次元画像変換行列 Τ η による変換処理は、 入力ビデオ信号について空 間的画像変換行列 Τ。 によって画像変換演算を実行することにより、 入力ビデオ 信号によって表されるオブジェク ト V I の映像が、 スク リ ーン 1 1 上の x、 y平 面上の位置から z軸上を奥行き方向に平行移動した後回転することにより、 ス ク リーン 1 1 の後方位置において奥行き情報をもつ変換映像になったことを意味す る。
この中間変換映像 V 2 は、 左目 1 2 Lからの視差を表わす画像変換をするため に、 次式
1 0 0
0 0 0
L = ( 2 ) 0 1 0
Δ χ 0 1 によって表される + Δ χ移動行列 Lによる変換処理をされ、 これにより中間変換 映像 V 3 Lが得られる。
この + Δ χ移動行列 Lは、 中間変換映像 V 2 を X軸に沿う方向に平行に、 距離 + Δ Xだけ移動させる変換演算を実行することを意味し、 かく して左目 1 2 しが ζ蚰上にあるものと考えて当該左目 1 2 Lの位置 1 2 L Xから中間変換映像 V 3 Lを見る状態になったことを意味する。
次に中間変換映像 V L 3 は次式
1 0 0
0 0 0
Ρ η = ( 3 )
0 0 Ρ
0 0 1 によって表される透視行列 Ρ。 を用いて変換処理されることにより、 X及び y軸 上にあるスク リ ーン 1 1上に中間変換映像 V 4 Lを求める。
この透視行列 P n による変換処理は、 中間変換映像 V 3 Lを左目 1 2 Lの位置 1 2 L Xから見たとき、 中間変換映像 V 3 Lの透視映像をスク リ ーン 1 1 上に透 視することにより見える映像は中間変換映像 V 4 Lであることを意味する。 続いてこの中間変換映像 V 4 Lは、 次式 1 0 0
0 0 0
L ( 4 )
0 1 0
厶 x 0 1 によって表される - Δ χ移動行列 L—'によって変換処理されることにより、 最終 変換映像 V 5 Lに変換される。
この一 Δ X移動行列 L—1は、 中間変換映像 V 4 Lを X軸方向に平行に距離一 Δ Xだけ平行移動させるもので、 かく して最終変換映像 V 5 Lは、 z軸を左目 1 2 L及び右目 1 2 Rの中間位置と して、 左目 1 2 Lの位置を当該中心位置から左側 に Δ χだけ移動させることにより、 左目 1 2 Lの本来の位置から中間変換映像 V 2を見たときスク リ ーン 1 1上に透視された映像 (図 4 ( Β 1 ) ) と して求めた ことになる。
.この左目用映像 1 4 Lは、 中間変換映像 V 2が奥行きをもっていることにより 、 手前の映像部分が大き く見えると共に、 奥行き方向に後方の映像部分が小さ く 見えるような透視映像を表わすことになる。
( 1 ) 式において使用されている変換パラメ一夕 r , ,〜 r 3 3は、 x、 y及び z 座標の 3次元空間において入力映像 V 1 を回転させるためのパラメ 一夕であり、 また変換パラメータ 、 ί y 及び £ z は、 x、 y及び z方向に平行移動させる ためのパラメ一夕であり、 変換パラメータ S t, は入力映像 V 1 を 2次元的に拡大 又は縮小するためのパラメータである。
また ( 2 ) 式及び ( 4 ) 式において使用されている変換パラメ 一夕 Δ X、 一 Δ Xは、 中間変換映像を X蚰方向に距離十 Δ X及び一 Δ Xだけ平行移動させるため のパラメ 一夕である。
さ らに ( 3 ) 式において用いられている変換パラ メ 一夕 P 2 は、 中間変換映像 V 3 Lをスク リ ーン 1 1上に透視する際に、 遠近法を適用するために用いられる パースぺクティブ値であり、 P 2 = ( 5 )
16 の値が基準値となるような値に設定されている (このことは、 オペレータの視点 が z座標の— 1 6の位置にあることを意味している。
図 4 (A 1 ) に示す左目用映像 1 4 Lについての変換処理をまとめれば、 次の ような演算式によって表わすことができる。
すなわち、 入力ビデオ信号の映像であるオブジェク ト V 1に基づいて、 左目用 の最終変換映像 V 5 Lを得るまでに、 オブジェク 卜 V 1 を中間変換映像 V 2に変 換する第 1の変換処理ステップと、 当該中間変換映像 V 2に基づいて左目用映像 1 4 Lを求める変換処理をする第 2の変換処理ステップとに着目すれば、 左目用 の変換行列 T L は、
T = T。 P 6 ) のように、 ( 1 ) 式による変換行列 T„ と、 その後の変換処理を表わす変換行列 T L との乗算式と して表される。
こ こで第 2の変換処理ステップで用いる変換行列 PL は、
P L = L P 0 L …… ( 7 ) に示すような透視変換行列であり、 ( 7 ) 式に ( 2 ) 式、 ( 3 ) 式及び ( 4 ) 式 を代入することにより、 1 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 1 0 0
P L =
0 1 0 0 0 P 0 0 1 0
Δ x 0 1 0 0 1 Δ x 0 0 1
Figure imgf000011_0001
と表わすことができる。
そこで、 ( 1 ) 式及び ( 8 ) 式を ( 6 ) 式に代入すれば、 左目用の変換行列 T
L は、 r r 1 1 r i r 1 3 0 1 0 0 0 〕 r 2 1 r 2 r 2 3 0 0 1 0 0
T L =
r a l r a - r 3 3 0 - P Δ x 0 0 P
£ , J2 y SL S 0 、 0 0 0 1 J
「 r i i 一 r i . P 厶 X r i 0 r 1 3 P ヽ r 2 1 - r 2 l P 厶 X r 2 0 r 2 3 P
Γ 】 3 — r 3; P 厶 X r a 2 0 r 3 3 P
χ - P Δ X 0 £ P + s。 j
( 9 ) のように表わすことができる。
と ころで、 ( 9 ) 式によって表される変換行列 T L は、 実際上入力ビデオ信号 が映像データがフ レームメモリ に取り込まれた後、 行列を構成するパラメ 一夕に よって変換処理を実行することになるが、 当該フ レームメモリから読み出される データ及びスク リ ーン 1 1上に表示される映像のデータは、 2次元データである ので、 ( 9 ) 式の第 3行目及び第 3列目のパラメータは、 2次元的なア ドレスを 演算する際には必要がない。
従って ( 9 ) 式から第 3行目及び第 3列目のパラメ一夕を除いた次式
T (10)
Figure imgf000012_0001
によって表される変換行列 T L33 力く、 フ レームメモリ上の位置ベク ト ル 〔 x、 y 〕 をスク リ ーン 1 1 上の位置べク トル 〔XS 、 Y s 〕 に変換するための行列と し て.用いられることになる。
こ こで、 ( 1 ) 式〜 ( 4 ) 式に示すような 4行 4列の変換行列は、 回転変換と 拡大、 縮小との違う次元の変換を同じ座標系で表しているので、 4 X 4 の行列と なり、 一般的にはこれを同次座標系 (Homogeneous Coordinate) と称している。 そ こで、 フ レームメ モリ上の 2次元的な位置べク トル 〔 X、 y〕 及びスク リ ー ン 1 1上の 2次元的な位置べク トル 〔XS 、 Y s 〕 を同次座標系で表わすと、 〔 x、 y〕 は 〔x、 y、 H] と表わすことができ、 〔 X s 、 Y s 〕 は 〔 X s 、 Y s 、 1 〕 と表わすことができる。 Hは、 遠近法により変形された映像拡大率又は縮 小率を表わす値であって、 z軸方向の位置べク トルの値と比例している。
フ レームメ モ リ上の位置べク トル 〔x、 y、 H〕 に対して変換行列 TL33 を作 用させると、 スク リ ーン 1 1上のべク トル 〔X s 、 Y s 、 1 j を求めることがで き、 求めたスク リ ーン 1 1上のべク トルは、
〔X Y 1 ) C y H] T (11) で定義される。
と ころが、 この実施例においては、 空間的画像変換を行う ときに、 入力ビデオ データをフ レームメモ リから読み出す時に 3次元変換するように、 ラスタ スキヤ ンに対応してシーケンシャルに供給されるスク リ ーンァ ドレスについてフ レーム メ モ リ上のア ドレスを指定する。
すなわち、 図 5に示すように、
〔x y H] = 〔X Y 1〕 T (12) のような逆演算を実行することにより、 スク リーン 1 1上の位置ベク トル 〔XS 、 Ys 、 1〕 がラスタスキャンに応じて順次指定されて行く と き、 これに応じて 対応するフ レームメモ リ上の位置ベク トル 〔x、 y、 H〕 を指定して行く よう に す.る。 かく してこの位置べク トル 〔x、 y、 H〕 を 2次元的な読出しア ドレスと して、 フ レームメ モ リ及びキーメ モ リ に供袷するこ とにより、 空間的画像変換を した 2次元的なビデオデータ及びキ一信号を得ることができる。
(12) 式の逆変換行列 TL331は、 次のようにして求めることができる。 まず、 変換行列 TL33 の各要素を、 r 1 1 - r 1 3 P 厶 x r , 3 Ρ χ
T r 2 1 - r 23 P Δ χ Γ 23 Ρ ζ
J2 χ - £ 2 Ρ 2 厶 χ £ 2 Ρ S
Figure imgf000013_0001
のよう に、 ノ 、。ラ メ - —夕 a i 1〜 33 次式
' a 3 】 3 一 r b , , b
Tし 3 a 3 b b b …… (14)
、 3 31 3 3 b b のよ う に、 パラ メ ータ ゎ 〜 によって表わす。
(14) 式を、 (12) 式に代入すれば、 r b , , b b
〔x y H〕 = [Xs Y s 1 b 2 b 22 b 23 …… (15) b a b b と表わすことができ、 これを展開すれば、
〔 X y H〕 [ b , X s + b . Y s + b ,
b + b Y "t" b
b 1 X + b 32 Y + b (16) なる。 従って (16) 式の関係から、 次式.
- b , , X s + b Y s + b (17) y , X + b Y 十 b (18)
H = b X s + b Y s + b (19) によって、 フ レームメモリ上の位置べク トル (x、 y、 H ) の各要素の値が演算 される。
と ころが、 このフ レームメモリ上の位置ベク トル 〔x、 y、 H〕 は、 同次座標 系のべク 卜ルであるので、 これを 2次元座標系に戻すには、 パラメータ Hで座標 値 x、 yを正規化すれば良い。
よって、 フ レームメモリ上のア ドレス 〔XLm、 YLm〕 は、
X
X
H b , , X s + b , z Y s + b
(20) b X + b Y 十 b y
Y Ln,=
H b X + b Y s + b
(21) b 3 , X s + b 32 Y s + b と して求めることができる。
かく してスク リ ーン用ア ドレス (Xs 、 Y s ) に対応する画素についてのデ一 夕を、 フ レームメモリ及びキー情報メモリからア ドレスデータ X Lm、 YLmを用い てシ一ケンシャルに読み出すことができる。
(20) 式及び (21) 式において、 b 〜 b 33の値は、 (14) 式から、
3 S 3.
b (22)
W 3 3 12
b (23) w
3 3 + a 12 a
b (24)
W ,
b (25) w
+ a
b (26) w
3 i 3 i 3 ] i a
b (27)
W ,
3 3 + 3
b (28) w
3 3
b (29)
W
a 1 a + a a
b (30)
W , になる。 ただし、 である。 ここで、 a u a 33の値は (13) 式の関係から、
a ] i = r i , - r i 3 P 2 Δ χ , a 1 2 = r 1 2 , a ] 3 = r 1 3 P 2 (32) a 2 i = r 2 i - r 23 P z Δ x , a 22 = r 22. a 23 ^ r 23 P i (33) a 31= J2 x - je 2 P 2 Δ χ , a 32 = y . a 33 = ^ z P , + S 0
-" (34) であるから、 これを (22) 式〜 (31) 式に代入することにより、
- Si y 1- 23 P , + r 22 ( £ z P , + S 0 )
b (35)
W ,
£ y r 1 3 P z + r 12 ( £ , P z + S。 )
b 1 2 = (36)
W,
一 Γ 22 Γ 13 Ρ 2 + Γ 12 Γ 23 Ρ ζ
b , 3 = (37)
W,
( ,一 ·δ ζΡ ,厶 x ) r 23 Ρ 2—( r 2,— r 23Ρ ζ厶 χ ) ( £ Ζ Ρ ζ + S 0 ) b 2 1 =
W,
(38) - ( J2 x- J2 z P ζΔ x ) r P z+ C r u— r 1 3 P ,厶 x ) ( je , F 2 + S 0) b 22 = :
W,
…… (39)
( r 2 1— r 23 P z厶 x ) r 1 3 P z— ( r 门— Γ 1 3 Ρ Ζ Δ Χ ) r 23 P z b 23 =
W,
…"- (40)
- r 22 ( J8 x - & t P z A x ) + ( r 2 , - r 23 P 2 Δ x ) & y b
W ,
…… (41) r , 2 ( J2 一 J2 , P A x ) - ( r , , - r , 3 P , A x ) &、 b a 2 =
. W,
…… (42)
- r 2 ( r 2 1 - r 2 3 P A x ) + ( r i , - r ] 3 P 2 Δ x ) r 22 b 33
W ,
…… (43)
W, = - r 2 2 ( & x - £ z P : Δ x r 13 P
+ ( r 2 ! - r 23 P , Δ x ) S> , r , a P z
+ r , 2 & — P Δ x ) r 23 P:
一 ( r , , - r , 3 P Δ x ) SL y r 23 P
— r 1 2 ( r 2 1— r 23 P z 厶 x ) ( , P 2 + S 0 )
+ ( r , , - r , 3 P Δ x ) r 22 ( P 2 + S 0 ) …… (44) のように、 ( 1 ) 式〜 ( 4 ) 式の変換行列において、 オペレータによって設定さ れるパラメ 一夕を用いて表わすことができる。
かく して、 (35) 式〜 (44) 式の値を (20) 式及び (21) 式に代入することに より、 フ レームメ モ リ に供給される読出しア ドレス (X Lm、 Y Lm) は、
X Lm = & , r 23 Ρ 2 + ( J0 s Ρ 2 + S ο ) } X
H
+ { ^ y r , 3 P z + r 1 2 ( j2 J P 2 + S o ) } Y
+ (一 Γ Γ P + Γ 1" I" 〕
(45)
Y Ll 〔 1 S, , - ΰ. z P z A x ) P
H
— ( r 21— r 23 P z 厶 x ) ( £ z P z + S o ) 1 X
+ ( - ( x - S.2 P 2 Δ x r P z
+ ( r n - r .s P , Δ χ ) ( , P z + S « ) 1 Y
+ { r ! - r P Δ x ) r 1 3 P 2
- ( r i ! - r , P z Δ x ) r 23 P , 1 〕
(46) と して与えられる。
ここで、
H = { ~ r ( L % - £ , P τ Δ χ ) + ( r 2 r 23 P , Δ x ) SL y } X + i r 1 2 ( 2 , — ·δ ζ Ρ ζ 厶 x ) — ( r , r , 3 P , Δ x ) Ά y } Y + { - r ( r - r P , Δ x ) + ( r - r P ! A x ) r
…… (47) である。
以上は、 図 4 ( A 1 ) において、 オブジェク ト V 1 に対する左目用画像変換演 算であるが、 右目用画像変換演算は図 4 ( A 2 ) に示すよ うに、 左目用画像変換 に対応するように実行される。
すなわち、 入力ビデオ信号の映像と してスク リ ー ン 1 1 上の x y平面の原点に 入力されたォブジェク 卜 V 1 は、 ( 1 ) 式によって表わされる 3次元変換行列 T によって Z軸上を奥行き方向に平行移動されると共に、 回転されて中間変換映 像 V 2 に変換される。
続いて中間変換映像 V 2 に対して右目の視差を表わす変換処理と して、 ( 4 ) 式によって表わされる一 Δ X移動行列 L—1によって X軸方向に平行に距離一 Δ X だけ平行移動することにより、 中間変換映像 V 2を中間変換映像 V 3 Rに変換す る。
.続いて、 中間変換映像 V 3 Rに対して ( 3 ) 式によって表わされる透視行列 P による変換がなされることにより、 中間変換映像 V 3 Rをスク リ ー ン 1 1 に透 視した中間変換映像 V 4 Rが得られる。 この中間変換映像 V 4 Rは、 右目 1 2 R が z軸上の位置 1 2 R Xにあるとき、 中間変換映像 V 2を右目 1 2 Rの視線方向 にスク リ ー ン 1 1 上に透視した映像 V 4 Rを得たことを意味している。
この中間変換映像 V 4 Rは、 ( 2 ) 式によって表わされる。 移動行列 Lによつ て変換されることにより、 X軸方向に距離 + Δ Xだけ平行移動して右目用の最終 変換映像 V 5 Rに変換される。
このときの右目用の最終変換映像 V 5 Rは、 右目 1 2 Rの位置を z軸上の中心 位置 (左目 1 2 L及び右目 1 2 Rの中心位置) から右方向に Δ Xだけずれた本来 の位置にある右目 1 2 Rによって、 中間変換映像 V 2 を見たとき、 当該中間変換 映像 V 2を右目 1 2 Rの本来の視線方向についてスク リ ーン 1 1上に透視した映 像であることを表している。
かく して右目用の最終変換映像 V 5 Rは、 図 4 ( B 2 ) に示すように、 奥行き をもつ映像と して、 手前の映像部分が大き く かつ奥行き方向の映像部分が小さ く なるような遠近法に基づく処理がなされた X y平面上の右目用映像 1 4 Rと して 目視することができることになる。
図 4 ( B 1 ) において、 入力ビデオ信号に基づく オブジェク 卜 V 1 を右目用の 最終変換映像 V 5 Rに変換する変換処理は、 左目用の変換処理と同様に、 次のよ うに実行される。
すなわち右目用の演算処理における変換行列 TR
TR = T P …… (48) のように、 オブジェク 卜 V 1 を z軸上に移動、 回転変換する 3次元変換行列 T„ と、 その後の右目用の変換処理を実行する変換行列 PR とに分けて考えることが でき、 当該右目用変換行列 P R は、
P R = L— 1 P。 L …… (49) のように、 ( 4 ) 式の一 Δ X平行移動行列 L—1と ( 3 ) 式の透視行列 P。 と、 ( 2 ) 式の + Δ X移動行列 Lとによって表される。
(49) 式に ( 4 ) 式、 ( 3 ) 式及び ( 2 ) 式を代入すれば、
1 0 0 0 0 0 1 0 0 0
0 0 0 1 0 0 0 1 0 0
P
0 1 0 0 0 P 0 0 1 0
Δ x 0 1 0 0 1 Δ x 0 0 1
1 0 0
0 0 0
(50)
P , 厶 x 0 P
0 0 1 のように展開できる。
そこで (48) 式に (50) 式を代入して展開すれば、 右目用の空間的画像変換を 行うための変換行列 P R は、 , ,
J
Figure imgf000022_0001
+ s。 J
(51) のように展開できる。
(51) 式において、 この場合も X y平面だけの有意情報を得るために、 (51) 式の 3行目及び 3列目のパラメ一夕を除いた次式、 r ! , + r , P 2 Δ x r r l 3P z
r r P 2 Δ x r 23 P z (52) J2 , + £ t P z Δ x £ P 2 + S によって表される右目用の変換行列 TR33 を用いる。
と ころで右目用の変換処理においても、 図 5について上述したようにフ レーム メ モ リ上の位置ベク トル 〔x y H〕 に対して変換行列 TR33 を作用させるこ とにより、 スク リーン 1 1上のべク トル 〔XS Y s 1 〕 を得ることになるの で、 スク リ ー ン 1 1上のベク トルは、
〔X Y 1 ] = C X y H〕 T (53) の^:によって定義される。
そ してビデオデータをフ レームメモリから読み出す際に用いる読出しア ドレス を求めるために、
X y H〕 〔X Y 1 〕 Tし (54) の逆演算を実行することにより、 スク リ ーン 1 1 上の位置べク トル 〔X S Y s 1 〕 がラスタスキャンに応じて指定されたと き、 対応するフ レームメ モ リ の位 置ベク トル 〔x y H〕 を指定し、 これにより空間的画像変換された 2次元の ビデオデータ及びキー信号を得るようにする。
ここで逆変換行列 TL33 1を求めるために、 r i I + r I a P z Δ x Γ 1 2 ! 3 P 2
丁 r 2 ! + r a s P z Δ x Γ 22 Γ 23 P z
J2 » + P I Δ x z P z 十 S
3 1 1 d l 3 l;
32 1 32 2 32: (55) a 3 1 d 3 2 33;
T R 33 1 …… (56)
Figure imgf000024_0001
のよ う に、 パラメータ a 〜 a 33及び b ~ b 33を、 変換行列 T R 33 及び逆変換 行列 T R33 一'の要素と して置いて、 パラメータ b , ,〜 b 33を求めれば、
332323 + 322333
b , , = (57)
W
332 d 一 3 】 23
b 1 2 = (58)
w
1 3 = (59)
w
33 l 3 23 — 32〗 3
b (60)
w + a
(61) w,
]3 23 = (62)
W, 一 3 S + a
(63) w a a
(64)
W
(65) w のようになる。
ここで
W — a a 1 a 1 + a 1 a a + a 1 a 1 a
一 3 I I a ~ a a a + a 1 1 3 22 3 33 (66) である。
(55) 式の関係からパラメ一夕 a , ,〜 a 33は、 a ii= ii + r i3P 2 Δ χ , a 12 = r , 2. a i3= r ,P (67)
3 21 = Γ 21 + 23 Ρ ι Δ X , 3 22 = Γ 22. 3 = Γ 23 P (68) a 3 1 = 1 , + 1 z P A x, a 32 = 1 y , a 33= 1 z P , + S 0
…… (69) であるから、 これを (57) 式〜 (66) 式に代入することにより、
Ά r 23P z + r 2 2 ( S. P , + S 0 )
b 】,= (70)
W
P z + r , 2 ( J2 Z P Z + S o )
(71)
W
— Γ 2 2 Γ 1 3 Ρ Ζ + l 2 Γ 2 3 P 2
b , 3 = …… (72)
W,
( J2 + ^ Z P ZA X) r 23 P z- ( r 21 + r 23 P Δ X ) ( J2 , P Z+ S 0) b 2 1 =
W,
…… (73)
— ( ^ ,+ J2 , P ,厶 x) r 1 3 P I+ ( r ,1+ r ,3 P 1A x) (je I P i + S o) b 22 =
W,
…… (74)
( r 21+ r 23F ,A x ) r , a P z - ( Γ + Γ , 3 Ρ Ζ Δ Χ) r 23P z b 2 3 =
W ,
…… (75) ( S. , + S. τ P t Δ x ) + ( r + r P Δ x ) £ b
W
(76) r , ( S. S.2 P 2 Δ x ) 一 ( + r , P Δ x ) S, b
W,
(77) 一 r ( r + P Δ x ) + ( + P Δ x )
W
(78)
W, = - r 22 ( J2 P Δ x ) r 13P 2
+ ( r . + r P Δ x ) £ y r P
+ r ( J2 + S. P , Δ x ) r 23 P ,
- ( r r 13P Δ x ) Ά y r 23 P ,
- r ( r , + r P , Δ x ) ( , P + S
+ ( r , , + r ,3P Δ x ) r 22 ( £ P + S (79) のように、 行列の各要素の値を、 ( 1 ) 式〜 ( 4 ) 式のようにオペレータが変換 条件と して設定するパラメ一夕によって表すことができる。
と ころで右目用の変換演算においても、 (15) 式〜 (21) 式について上述した と同様にして、 (54) 式に (56) 式の関係を代入することにより、 r b 12 b , 3、
〔 X y H〕 = [Xs Ys 1 〕 b b b …… (80) b b b のように、 フ レームメモリ上の位置べク トノレ 〔x y H〕 を、 スク リ ーン 1 1 上の位置べク トル 〔XS Y s 1〕 と、 パラメータ b , ,~ b 33とによって表す ことができ、 これを次式、
〔 X y H〕 C b 1 , X S + b , 2 Y S + b 1 3
2 1 X S + b 22 Y S + b 23
b 3 I X S + b 32 Y S + b 33 ] (81) のように展開することにより、 フ レームメモリの位置べク トル 〔x y H〕 を
x = b Xs + b 12Ys + b 13 …… (82) y = b 2.Xs + b 22Ys + b 23 …… (83)
H = b 31Xs + b Y s + b ······ (84) と して求めることができる。
そ してフ レームメ モ リ上の位置べク トル 〔x y H〕 が同次座標系のべク 卜 ルであることに基づいて、 これをパラメータ Hで正規化することによりフ レーム メモリ上のア ドレス (XRm YRm) と して、
X
XRm= ——
H b! , X S + b ! 2 Υ S + b ! 3
…… (85) b 3 1 X S + b 32 Y S + b 33 y
Y
H b 2 i X s + b 22 Y s + b
(86) b a i X s + b 32 Y s + b と して求めることができる。
そこで、 (85) 式及び (86) 式に (70) 式〜 (79) 式を代入することにより フ レームメ モ リ に対する読出しア ドレス (XRm YRm) を、
XRm = C { - » r 23 P , + ( z P 2 + S o ) X
H
+ I S. y r 1 3 P z + r 12 ( J2 Z P Z + S 0 ) ) Y
+ ( - r 22 r , a P z + r , 2 r 23 P z ) 〕
(87)
Y R, 〔 { S. x + £ 2 P z Δ x ) r P
H
一 ( r 2 , + r 2 S P , Δ x ) ( £ τ P , + S„ ) I X
+ { - ( £ x + & 2 P z Δ x r , 3 P 2
+ ( l- , , + r , a P z A x ) ( & 2 P τ + S u ) } Y
+ i r 2 , + r 23 P 2 Δ x ) r 13 P z
- ( r , , + r , 3 P , Δ x ) r 23 P , } J
(88) の値に指定すれば良いことになる
こ こで、 H = { - r 2 ( J2 x + S. z P z A x ) + ( r 21+ r 23P z A x ) - £ y } X + { r , 2 ( Jg x + S, z P 2 A x ) - ( r , , + r , 3 P , Δ x ) - Ά y \ Y + ( - 1- 1 2 - ( r 2 1 + r 23 P 2 Δ x ) + ( r , , + r , 3 P , Δ x ) r 2 2 i
······ (89) である。
以上の立体映像の生成方法によれば、 入力ビデオ信号によって表されるォブジ ェク 卜 V I を、 奥行き方向に移動、 回転させると共に、 Δ χ及び— Δ χ方向に移 動させることにより得た中間変換映像 V 3 L及び V 3 Rを、 それぞれ左目 1 2 L 及び右目 1 2 Rの視線に沿う方向においてスク リーン 1 1上に透視するこ とによ り、 左目 1 2 L及び右目 1 2 R間の視差をもってスク リ ー ン 1 1上に投射すべき 最終変換映像 V 5 L及び V 5 Rを求めることができ、 これにより左目 1 2 L及び 右 1 2 Rの視差に基づいて奥行きをもった物体の映像をスク リーン面上、 又は スク リ ー ン面の後方位置、 又はスク リー ン面の前方位置に生成させることができ る。
そ して立体映像の生成を、 ( 1 ) 式によって定義される 3次元変換行列 Τ 0 、 ( 2 ) 式によって定義される + Δ X移動行列 L、 ( 3 ) 式によって定義される透 視行列 P。 及び ( 4 ) 式によって定義される - Δ X移動行列 L 1を規定する各パ ラメ一夕をオペレータが入力することにより、 必要に応じて任意の位置に任意の 変換対応をもつ立体映像を容易に生成することができる。
かくするにっき、 設定できるパラメータと して、 変換映像の空間的な回転を表 すパラメ 一タ r , ,~ r 33や、 X軸、 y軸及び z軸方向の平行移動を表すパラメ一 夕 J2 , 、 S, y 及び J2 , や、 2次元的な拡大圧縮を表すパラメータ や、 遠近法 に基づく パースペクティブ値を表すパラメ ータ P z や、 左目 1 2 Lと右目 1 2 R との距離の半分の距離を示す値 Δ Xを必要に応じて任意に選択して指定できるこ とにより、 生成すべき映像を容易に特定できる。 また入力ビデオ信号の映像を表す映像データを書き込むと共に、 その読出しァ ドレスをスク リ ーン上のラスタスキャンァ ドレスに対応させて指定して行く こと ができるように したことにより、 スク リーン 1 1 上への左目用及び右目用の最終 変換映像 V 5 L及び V 5 Rを同時に生成させることができる。
( 2 ) 立体映像生成装置の第 1 の実施例
図 6 において、 2 1 は全体と して本発明による立体映像生成装置を示し、 入力 ビデオ信号 VA と、 入力キー信号 KA と、 入力バックグラウン ド信号 VBKとを、 左目用プロセッサ 2 2 L及び右目用プロセッサ 2 2 Rに受けて、 それぞれ左目用 及び右目用の最終変換映像 V 5 L及び V 5 Rを表す左目用及び右目用立体映像出 力信号 V L。u t 及び V R。U, を左目用及び右目用プロジェクタ 2 3 L及び 2 3 R に供給する。
かく して左目用及び右目用プロジェクタ 2 3 L及び 2 3 Rは、 スク リ ーン 2 4 上 こそれぞれ左目用及び右目用最終変換映像 V 5 L及び V 5 Rをそれぞれ映出さ せる。
左目用プロセッサ 2 2 L及び右目用プロセッサ 2 2 R、 並びに左目用及び右目 用プロジェクタ 2 3 L及び 2 3 Rは、 バス 3 1 を介して中央処理ュニッ ト ( C P U) 3 2 によって制御される。
C P U 3 2 は R OM構成のプログラムメモリ 3 3のプログラムに従って、 RA M構成のヮ一キングメモ リ 3 4を利用しながら、 立体映像生成装置 2 1 の各部の 処理動作を実行する。
C P U 3 2 は、 書込みア ドレス発生回路 3 5からフ レームメ モ リ上の位置べク トル 〔 x y〕 を表す書込みア ドレス信号 S X及び S yを発生させ、 この書込み ア ドレス信号 S x及び S yによって左目用及び右目用プロセッサ 2 2 L及び 2 2 Rのフ レームメモリ 3 6 L及び 3 6 Rのメモリ位置を順次指定して行く ことによ り、 入力ビデオ信号 VA を左目用及び右目用ソースビデオデ一夕と して互いに同 期しながら記憶させる。 害込みァ ドレス信号 S x及び S yは左目用及び右目用プロセッサ 2 2 L及び 2 2 Rのキー信号メモリ 3 7 L及び 3 7 Rにも与えられ、 これにより入力ビデオ信 号 VA と同期しながらフ レームメ モ リ上の位置べク トル 〔 x y〕 についてのキ —デ一夕をキー信号メモリ 3 7 L及び 3 7 Rに記憶させる。
C P U 3 2 は左目用及び右目用プロセッサ 2 2 L及び 2 2 Rに共通に設けられ ているスク リ ーンァ ドレス発生回路 3 8を駆動することにより、 スク リーン 2 4 上の位置べク トル 〔XS Y s 〕 を内蔵の基準信号発生回路から発生された基準 信号に基づいて発生させ、 このスク リ ー ンァ ドレス信号 S X s 及び S Y s を左目 用及び右目用プロセッサ 2 2 L及び 2 2 Rの読出しァ ドレス発生回路 3 9 L及び 3 9 Rに供給する。
読出しァ ドレス発生回路 3 9 L及び 3 9 Rは、 C P U 3 2の制御の下に、 それ ぞれ (45) 式、 (46) 式及び (87) 式、 (88) 式によって特定された読出しァ ド レス (Xtm YLm) 及び (XRm Y Rm) を表す読出しア ドレス信号 S XLm、 S Y 及び S XRm、 S YRmをフ レームメ モ リ 3 6 L及び 3 6 Rと、 キーデ一夕 メ モ リ 3 7 L並びに 3 7 Rとに供給する。
かく して左目用及び右目用プロセッサ 2 2 L及び 2 2 Rのフ レームメモリ 3 6 L及び 3 6 Rから読み出された映像データ V L , 及び V R , はビデオ信号補間回 路 4 0 L及び 4 0 Rにおいて補間処理された後、 読出し映像信号 V L 及び V R と してコ ンバイナ 4 1 L及び 4 1 Rに与えられる。
これと同時に左目用及び右目用プロセッサ 2 2 L及び 2 2 Rのキ一信号メ モ リ 3 7及び 3 7 Rから読み出されたキーデータ K L 及び K R , はキー信号補間回 路 4 2 L及び 4 2 Rにおいて補間処理された後、 読出しキー信号 K L 2 及び K R と してコ ンパイナ 4 1 L及び 4 1 Rに与えられる。
コンバイナ 4 1 L及び 4 1 Rは、 C P U 3 2の制御の下に、 読出しキー信号 K L 2 及び K R2 によって、 読出し映像信号 V L 2 及び V R 2 と、 入力バック グラ ゥ ン ド信号 VBKとの間のキーイ ング処理を実行し、 その結果次式、 V L K L 2 V L 2 + ( 1 - K L ) V BK (90)
V R U , = K R 2 V R 2 + ( 1 - K R 2 ) V BK ··· ··· (91) によって表される立体映像出力信号 V L。u t 及び V R。u t を、 左目用及び右目用 プロセッサ 2 2 L及び 2 2 Rの出力と して、 左目用及び右目用プロジェクタ 2 3 L及び 2 3 Rに供給する。
次に、 この立体映像生成装置 2 1 の動作を説明する。
まず、 オペレータは、 本発明の演算に予め必要とされる視差距離 Δ X及びパー スぺクティ ブ値 P z の値を、 コ ン ト ロールパネル 5 1 を介して入力する。 C P U 3 2 は、 オペレー夕によって新たに設定された視差距離 Δ X及びパースぺクティ ブ値 P , を、 R A M 3 4 に記憶する。 なお、 C P U 3 2 は、 オペレータからの入 力がないことを検出すると、 予め R A M 3 4 に記憶されている基準値 (Δ χ = 3 . .2 5 c m, Ρ Ί = - 1 / 1 6 ) を使用する。
次に、 オペレータは、 コ ン ト ロールパネル 5 1 に設けられた 3次元ポイ ンティ ングデバイス等を操作することによって、 ソースビデオ信号に対して 3次元の画 像変換操作を指令する。 オペレータによって、 ソースビデオ信号に対して何らか の 3次元変換の処理が指示されると、 C P U 3 2 は、 オペレータが指定した 3次 元変換を表す行列式 Τ。 の各パラメ 一夕である、 「 〜!" 、 £ t 及び S N を受け取ると共に、 また、 同時に R A M 3 4 に記憶された視差距離 Δ X 及びパースペクティブ値 P: とを受け取る。 次に、 C P U 3 2 は、 受け取ったパ ラ メ一夕 i- , ,~ r 33 ί , i y i t S。 Δ χ及び とを使用 して (35〉 式〜 (43) 式で示されるパラメータ b , , b 33の値、 及び、 (70) 式〜 (78) 式 で示されるパラメ一夕 b , !~ b 33の値を演算する。 C P U 3 2 は、 演算したパラ メ ータ b , ,~ b 33の値を、 読出しア ドレス発生回路 3 9 Lに供給し、 同時に、 演 算したパラメータ 〜 ^の値を読出しァ ドレス発生回路 3 9 Rに供給する。 読出しァ ドレス発生回路 3 9 Lは、 C P U 3 2から供袷されたパラ メ ータ b M 〜 b 33と、 スク リ ーンァ ドレス発生回路 3 8から供給されたスク リーンァ ドレス CX s 、 Y s ) とに基づいて、 (20) 式において示されるような左目用読出しァ ドレス ( X Lm、 Y Lm) を生成する。 この読出しァ ドレス発生回路 3 9 Lは、 生成 された左目用読出しア ドレス (X Lm、 Y Lm) を、 フ レームメ モ リ 3 6 L及びキ一 信号メ モ リ 3 7 Lに供給する。
また、 同時に、 読出しァ ドレス発生回路 3 9 Rは、 C P U 3 2から供給された パラ メ 一夕 b l l〜b 33と、 スク リ ー ンア ドレス発生回路 3 8から供給されるス ク リ ーンア ドレス (X Lm、 Y Lm) とに基づいて、 (85) 式において示されるよ うな 右目用読出しア ドレス (XRm、 YRm) を生成する。 読出しァ ドレス発生回路 3 9 Rは、 生成された右目用読出しア ドレス (XRm、 Y Rm) を、 フ レームメ モ リ 3 6 R及びキー信号メモリ 3 7 Rに供袷する。
このよう に、 フ レームメ モ リ 3 6 Lから読み出された変換ビデオ信号 V L , は 、 図 4 ( A 1 ) に示されるように、 空間的変換行列 T。 によって 3次元空間上に 変換されたビデオ信号 V 2を、 左目の視点位置 1 2 Lから見たときに、 ス ク リ ー ン面に投影されるビデオ信号 V 5 Lを示し、 フ レームメ モ リ 3 6 Rから読み出さ れた変換ビデオ信号 V R , は、 図 4 ( A 2 ) に示されるように、 空間的変換行列 T 0 によって 3次元空間上に変換されたビデオ信号 V 2 を、 右目の視点位置 1 2 Rから見たときに、 スク リ ーン面に投影される ビデオ信号 V 5 Rを示している。 フ レームメ モ リ 3 6 L及び 3 6 Rから読み出された変換ビデオ信号 V L , 及び V R , は、 ビデオ信号補間回路 4 0 L及び 4 0 Rによって、 周辺画素によって補 間処理されて、 変換ビデオ信号 V L 2 及び V R2 と してそれぞれ出力される。 同 様に、 キー信号メモ リ 3 7 L及び 3 7 Rから読み出された変換キ一信号 K L , 及 び K R , は、 キー信号補間回路 4 2 L及び 4 2 Rによって、 周辺画素によって補 間処理されて、 変換キー信号 K L 2 及び K R2 と してそれぞれ出力される。 このよう に して、 C P U 3 2は、 オペレータがコ ン ト ロールパネル 5 1 から入 力した入力情報をィ ンタ一フェース 5 2を介して取り込むことにより、 ォペレ一 夕が指定した画像変換処理内容によって入力ビデオ信号 V A 及び入力キ一信号 K A を同時に変換処理してなる左目用及び右目用立体映像出力信号 V L。u t 及び V R ou , を左目用及び右目用プロセッサ 2 2 L及び 2 2 Rから左目用及び右目用プ ロジェクタ 2 3 L及び 2 3 Rに与える。
ここで、 オペレータがコ ン トロールパネル 5 1 からパラメ 一夕 1- , ,〜 r 33を設 定すれば、 C P U 3 2 は、 銃出しァ ドレス発生回路 3 9 L及び 3 9 Rにおいて 3 次元変換行列 P。 について、 オブジェク ト V 1 をどの程度空間的に回転変換すベ きであるかを指定する。
また、 オペレータがコン トロールパネル 5 1 からパラメータ J2 x 、 £ y 及び Jg , を設定すれば、 C P U 3 2 は読出しァ ドレス発生回路 3 9 L及び 3 9 Rにおい て 3次元変換行列 P。 について、 オブジェク ト V I を x$由方向、 y軸方向及び z 軸方向にどの程度平行移動すべきであるかを指定する。
さ らに、 オペレータがコン トロールパネル 5 1 からパラメ一夕 S。 を入力すれ ば、 C P U 3 2 は、 読出しア ドレス発生回路 3 9 L及び 3 9 Rにおいて、 3次元 変:換行列 P。 について、 X y平面上においてオブジェク ト V 1 をどの程度拡大又 は縮小すべきであるかを指定する。
さ らに、 オペレータがコン トロールパネル 5 1 からパラメ一夕 Δ χを入力すれ ば、 C P U 3 2 は、 読出しァ ドレス発生回路 3 9 L及び 3 9 Rにおいて、 + Δ χ 移動行列 L及び一 Δ X移動行列 L—1について、 左目 1 2 L及び右目 1 2 R間の距 離の半分の距離厶 Xをどの程度にすべきであるか (すなわち左目 1 2 L及び右目 1 2 R間の視差をどの程度にすべきか) を規定するような読出しァ ドレス信号 S X Lm、 S Y Lm及び S XRm、 S YRmを発生させる。
この実施例の場合、 パラ メ 一夕 P , については、 コ ン ト ロールパネル 5 1 から 常時 1 ノ 16の値が常時入力されるように構成されており、 かく して C P U 3 2 は 読出しァ ドレス発生回路 3 9 L及び 3 9 Rにおいて透視行列 P n について、 遠近 法に基づく パースぺクティブ値を一定値に規定するような読出しァ ドレス信号を 送出させる。
かく してオペレータは、 コ ン トロールパネル 5 1 から左目 1 2 L及び右目 1 2 Rの視差を規定するパラメータ Δ xを所望の値に選定できることにより、 左目用 及び右目用プロジェクタ 2 3 L及び 2 3 Rによってスク リ ーン 2 4上に投射され る最終変換映像 V 5 L及び V 5 Rでなる左目用映像 1 4 L及び右目用映像 1 4 R のずれ量を好みに応じて容易に適切な値に設定することができる。
かく するにっき、 左目用及び右目用プロセッサ 2 2 L及び 2 2 Rの読出しア ド レス発生回路 3 9 L及び 3 9 Rの処理は、 共通のスク リーンア ドレス信号 S X s 及び S Y S に基づいて同時に実行するようになされているので、 左目用及び右目 用立体映像出力信号 V L。u t 及び V R。u t が時間的に同時に発生され、 従って一 段と画質が良好な立体映像を生成することができる。
上述の構成によれば、 左目用透視変換行列 P L と右目用透視変換行列とを演算 に用いることによって、 ソースビデオ信号 V I を、 3次元空間において移動及び 回転させることによって空間的な位置に変換された変換ビデオ信号を、 左目の位 置 1 2 L及び右目の位置 1 2 尺から、 それぞれ X y平面に投影した、 左目用のビ デ.ォ信号と右目用のビデオ信号とを演算しているので、 左目と右目の視差に応じ た適切なビデオ信号を生成することができる。
また、 変換されたビデオ信号 V 2の空間的位置に応じた視差を有した左目用ビ デォ信号と右目用ビデオ信号を生成することができる。 言い換えると、 ソースビ デォ信号が、 ス ク リーンに対して 3次元的に奥方向 ( + z方向) の位置に変換さ れるほど、 左目用ビデオ信号が左方向にずれると共に、 右目用ビデオ信号が右方 向にずれるように演算される。 また、 ソースビデオ信号が、 逆に、 スク リ ーンに 対して手前方向 (― z方向) の位置に変換されるほど、 左目用のビデオ信号が右 方向にずれると共に、 右目用のビデオ信号が左にずれるように演算される。 よつ て、 容易に立体感のある映像信号を生成することができる。
( 3 ) 立体映像生成装置の第 2の実施例
図 7 は、 本発明の第 2の実施例を示すもので、 この場合立体映像生成装置 6 1 は図 6 との対応部分に同一符号を付して示すように、 左目用及び右目用プロジェ ク タ 6 2 L及び 6 2 Rから得られる左目用及び右目用立体映像出力信号 V L。U T 及び V R。u t を左目用及び右目用プロジェクタ 2 3 L及び 2 3 Rにそれぞれ出力 すると共に、 例えばビデオテープレコーダでなる左目用及び右目用映像信号記録 装置 6 3 L及び 6 3 Rに記録するようになされている。
左目用及び右目用映像信号記録装置 6 3 及び 6 7 1¾は、 C P U 3 2の制御の 下に、 記録した映像信号を再生して左目用及び右目用帰還映像信号 V B L及び V B Rと してそれぞれ切換スィツチ 6 4 L及び 6 4 Rの切換入力端 W 2 を通じて左 目用及び右目用プロセッサ 6 2 L及び 6 2 Rのフ レームメモリ 3 6 L及び 3 6 R に対する映像入力信号 V L。 及び V R。 と して帰還する。
この実施例の場合、 入力ビデオ信号 V A は、 C P U 3 2の制御の下に切換動作 する左目用及び右目用切換スィッチ 6 4 L及び 6 4 1¾の第 1 の切換入力端 W 1 を 通じて入力映像信号 W L。 及び W R。 と してフ レームメモ リ 3 6 L及び 3 6 Rに 入力される。
、く して左目用及び右目用切換スィ ッチ 6 4 L及び 6 4 Rを通じて入力された 入力映像信号 V L 0 及び V R。 は、 左目用及び右目用プロセッサ 6 2 L及び 6 2 Rのキー信号生成回路 6 5 L及び 6 5 Rにも入力され、 かく して C P U 3 2 の制 御の下に、 キー信号生成動作を実行するキー信号生成回路 6 5 L及び 6 5 尺から 、 入力映像信号 V L。 及び V Rn と同期してキーイ ング処理をすることができる 入力キー信号 K L。 及び K R Q をキー信号メモリ 3 7 L及び 3 7 Rに入力する。 図 7 の構成において、 図 8 ( A ) 及び図 8 ( B ) に示すように、 左目用及び右 目用プロジェク タ 2 3 L及び 2 3 Rからスク リーン 2 4 に投射される映像のうち 、 スク リ ー ン 2 4 の一部 (例えば左側部分) に第 1 の左目用及び右目用映像 7 0 L及び 7 O Rを投射すると共に、 スク リ ーン 2 4の他の部分 (例えば右側部分) に第 2 の左目用及び右目用映像 7 2 L及び 7 2 Rを投射する。
第 1 の左目用及び右目用映像 7 0 L及び 7 0 Rは、 入力ビデオ信号 V A によつ て表される映像を、 図 4 ( A 1 ) 〜図 4 ( B 2 ) について上述した手法で画像変 換処理をする。 この実施例の場合、 第 1 の左目用及び右目用映像 7 0 L及び 7 0 Rは、 左目 1 2 L及び右目 1 2 Rの視線がスク リ ー ン 2 4の後方位置に虚像を結像することに より、 入力ビデオ信号 VA に基づいて結像された立体映像 7 1 がオペレータによ つて知覚される。
これに対してスク リーン 2 4 の右側部分に投射された第 2の左目用及び右目用 映像 7 2 L及び 7 2 Rは、 左目用及び右目用映像信号記録装置 6 3 L及び 6 3 R から再生された左目用及び右目用帰還映像信号 V 8 し及び 81¾にょって表され る映像を、 図 4 ( A 1 ) 〜図 4 ( B 2 ) によって上述した画像変換手法によって 画像変換処理された読出しア ドレス信号 S XLm、 S Y Lm及び S XRm、 S YRmによ つてビデオメモリ 3 6 L及び 3 6 Rから読み出して、 スク リ ーン 2 4上に投影し たものである。
図 8 ( A ) の場合、 第 2の左目用及び右目用映像 7 2 L及び 7 2 Rに対する左 目 1 2 L及び右目 1 2 Rの視線は、 スク リ ーン 2 4 の手前で交差するこ とによ り 、 スク リ ーン 2 4 の前方にある仮想空間に左目用及び右目用帰還映像信号 V B L 及び V B Rによって結像された立体映像 7 3がオペレータによって知覚される。 図 7の構成において、 左目用及び右目用切換スィ ッチ 6 4 L及び 6 4 Rは C P U 3 2の制御の下に、 1 フ レーム分の映像区間のうち左側の部分のタイ ミ ングで 入力ビデオ信号 VA を第 1 の切換入力端 W 1 を通して左目用及び右目用フ レーム メ モ リ 3 6 L及び 3 6 Rの左側メ モ リ領域に記憶される。
これに対して左目用及び右目用切換スィツチ 6 4 L及び 6 4 Rは、 1 フ レーム 期間のうち左側部分の期間において第 2の切換入力端 W 2側に切り換わることに より左目用及び右目用帰還映像信号 V B L及び V B Rをフ レームメ モ リ 3 6 L及 び 3 6 Rに記憶される。
かく してフ レームメモリ 3 6 L及び 3 6 Rには 2つのビデオデータを含む 1 フ レーム分の映像データが記憶されることになり、 当該映像データが C P U 3 2の 制御の下に、 铳出しァ ドレス発生回路 3 9 L及び 3 9 Rにおいて画像変換処理さ れた読出しァ ドレス信号を発生することにより、 第 1 の左目用及び右目用映像信 号 7 O L及び 7 O Rと第 2の左目用及び右目用映像 7 2 L及び 7 2 Rとを表す 1 フ レーム分の映像でなる左目用及び右目用立体映像出力信号 V L。u t 及び V R。U t をプロジェク タ 2 3 L及び 2 3 Rに送出する。
この画像変換処理された 1 フ レーム分の映像を表す左目用及び右目用立体映像 出力信号 V L。u l 及び VR。u t は左目用及び右目用記憶装置 6 3 L及び 6 3 Rに 記憶された後、 第 2の左目用及び右目用映像 7 2 L及び 7 2 Rの映像信号だけが 左目用及び右目用切換スィ ッチ 2 4 L及び 2 4 Rを通して入力映像信号 V L。 及 び V R。 と してフ レームメモリ 3 6 L及び 3 6 Rに記憶される。
こ こで入力ビデオ信号 V A に変化がなければ、 フ レームメ モ リ 3 6 L及び 3 6 Rに記億された 1 フ レーム分の映像データのうち、 第 2の左目用及び右目用映像 7 2 L及び 7 2 Rだけが、 読出しァ ドレス発生回路 3 9 L及び 3 9 Rにおける画 像変換処理によって変更される。
この状態において C P U 3 2の制御の下に読出しァ ドレス発生回路 3 9 L及び 3.9 Rが再度画像変換処理を実行すれば、 第 2の左目用及び右目用映像 7 2 L及 び 7 2 Rだけが、 再度図 4 (A 1 ) 〜図 4 ( B 2 ) の画像変換処理をされた後、 左目用及び右目用立体映像出力信号 V L。u t 及び V R。u t と してプロジェクタ 2 3 L及び 2 3 Rに出力される。
かく してプロジェクタ 2 3 L及び 2 3 Rはスク リ ーン 2 4上の右側部分に映出 される第 2の左目用及び右目用映像 7 2 L及び 7 2 Rが、 繰り返し画像変換処理 を受けていることにより、 当該帰還映像信号 V B L及び V B Rによって結像され る立体映像 7 3が順次各フ レームごとに変化して行く。 これに対して、 映像の内 容が変化しない入力ビデオ信号 VA に基づく第 1 の左目用及び右目用映像 7 0 L 及び 7 0 Rによって結像された立体映像 7 1 は、 変化しないで、 固定した映像と してオペレータによって知覚される。
以上のように図 7の構成によれば、 スク リーン 2 4上に生成された立体映像の うち、 所望の一部分の立体映像だけについて左目用及び右目用帰還映像信号 V B L及び V B Rと してフ レームメモリ 3 6 L及び 3 6 Rに帰還できるようにしたこ とにより、 立体映像のうちの一部が繰り返し実行される画像変換処理によって変 ィ匕して行く ような立体映像を容易に生成することができる。
( 4 ) 他の実施例
上述の実施例においては、 透視行列 P。 についての ( 3 ) 式のパラメ 一夕 P z を予め 1 / 16に固定することにより、 遠近法に基づく パースペクティブ値を変更 しないようにした場合について述べたが、 これに限らず、 当該パラメータ P z も コ ン トロールパネル 5 1 に対するオペレータの操作によって変更できるようにし ても上述の実施例の場合と同様の効果を得ることができる。
また上述の図 7の実施例においては、 スク リーン 2 4 に投射された映像の一部 についてフ レームメモリ 3 6 L及び 3 6 Rに帰還することにより .、 当該一部の映 像について繰り返し画像変換処理をするようにした場合について述べたが、 本発 明はこれに限らず、 スク リーン 2 4上に生成される映像の全部について繰り返し 画像変換処理をするようにしても良い。
また上述の実施例においては、 左目用及び右目用立体映像出力信号 V L。u t 及 び V R。u t を左目用及び右目用プロジェクタ 2 3 L及び 2 3 Rによってスク リ ー ン 2 4上に投影することにより立体映像を生成するようにした場合について述べ たが、 立体映像の表示手段と しては、 これに限らず、 例えばヘッ ドマウン ト型映 像装置のような他の型式の映像表示装置を適応しても良い。
また図 7の実施例においては、 左目用及び右目用映像信号記録装置と してビデ ォテープレコーダを適用した場合の実施例について述べたが、 映像信号記録装置 と してはこれに限らず、 ビデオディスク装置や半導体メモリ装置などの他の映像 記録装置を用いるようにしても良い。
産業上の利用可能性
本発明は、 2次元的な映像を表すビデオ信号を用いて、 立体映像を生成する立 体映像生成装置に適用できる。

Claims

請 求 の 範 囲
1. 視覚的に立体効果を得るためのビデオ信号を生成する立体映像生成装置にお いて、
ソースビデオ信号によって表される映像を仮想的な空間位置に変換するような画 像変換をするための空間的画像変換行列と、 当該変換された変換ビデオ信号によ つて表される変換映像をスク リ ーン面上に透視するための左目用透視変換行列と に基づいて、 上記ソ一スビデオ信号から左目用立体ビデオ信号を生成する左目用 プロセッサと、
上記空間的画像変換行列と、 上記変換ビデオ信号によって表される変換映像を スク リ ーン面上に透視するための右目用透視変換行列とに基づいて、 上記ソ一ス ビデオ信号から右目用立体ビデオ信号を生成する右目用プロセッサと、
を具えることを特徵とする立体映像生成装置。
2.. 請求の範囲第 1項の立体映像生成装置において、
上記左目用立体ビデオ信号は、 上記左目の空間位置を視点と して上記仮想的な 空間位置にある変換ビデオ信号の映像を見たときに、 上記スク リ ーン面上に透視 される 2次元的なビデオ信号であり、 上記右目用立体ビデオ信号は、 上記右目の 空間位置を視点と して上記仮想的な空間位置にある変換ビデオ信号の映像を見た と きに、 上記スク リ ーン面上に透視される 2次元的なビデオ信号である
ことを特徴とする立体映像生成装置。
3. 請求の範囲第 1 項の立体映像生成装置において、
ラ スタスキャンに対応したシーケ ンシャルなスク リ ーンァ ドレスを発生させる ためのスク リ ーンァ ドレス発生手段をさ らに具え、
上記左目用プロセッサから上記左目用立体ビデオ信号が出力されるタ イ ミ ング と、 上記右目用プロセッサから上記右目用立体ビデオ信号が出力されるタイ ミ ン グとが同じタイ ミ ングとなるよ う に、 上記スク リ ー ンア ドレスに応答して、 上記 左目用プロセッサ及び上記右目用プロセッサが同時に動作する ことを特徴とする立体映像生成装置。
4. 請求の範囲第 1 項の立体映像生成装置において、
上記空間的画像変換行列の各パラメ ータは、 オペレータの所望の画像変換操作 に基づいて各フ レーム毎に設定される値であり、
上記左目用透視変換行列及び上記右目用透視変換行列の各パラメ 一夕は、 オペ レー夕によって予め設定された値である
ことを特徴とする立体映像生成装置。
5. 請求の範囲第 1項の立体映像生成装置において、
上記空間的画像変換行列 T D は、 r i r 1 r 1 3 0
r r r 0
T„ =
r :I I r 3 2 Γ 3 3 0
S. z s によって定義され、
上記左目用透視変換行列 P L は
1 0 0
0 0 0
P L
P z Δ X 0 P
0 0 1 によって定義され、
上記右目用透視変換行列 P R は、 1 0 0 0
0 1 0 0
P R
P z Δ x 0 0 P
0 0 0 1 によって定義され、
ただし、 1- , ,〜 r 3 3は空間的な回転変換を表すパラメ ータであり は X軸方向の平行移動を表すパラメータであり、
y は y軸方向の平行移動を表すパラメータであり、
■2 z は z軸方向の平行移動を表すパラメ一夕であり、
S„ は 2次元的な拡大縮小を表すパラメータであり、
P z は予め設定されたパースペクティ ブ値であり、
Δ Xは右目と左目の距離の半分の距離を示す値である
ことを特徴とする立体映像生成装置。
6. 請求の範囲第 5項の立体映像生成装置において、
上記左目用透視変換行列 P L は、
P L P 0 L によって求められ、 上記右目用透視変換行列 P R
P R = L 一 1 L P によって求められ、 ただし、 行列 Lは、 1 0 0
0 0 0
L
0 1 0
Δ x 0 1 によって定義され、 行列 L—1は.
1 0 0 0
0 1 0 0
0 0 1 0
Δ X 0 0 1 によって定義され、 行列 Ρ ο は
1 0 0 0
0 1 0 0
Ρ
0 0 1 Ρ
0 0 0 1 によって定義される
ことを特徴とする立体映像生成装置。
7. 請求の範囲第 6項の立体映像生成装置において、
上記左目用プロセッサは、
上記ソースビデオ信号を一時的に記憶する第 1 の記憶手段と、
上記空間的画像変換行列と上記左目用透視変換行列とに基づいて演算され、 上 記第 1 の記憶手段に記憶されたビデオ信号を読み出すための読出しア ドレスを生 成する第 1の読出しア ドレス生成手段と
を有し、 上記右目用プロセッサは、
上記ソ一スビデオ信号を一時的に記憶する第 2 の記憶手段と、
上記空間的変換行列及び上記右目用透視変換行列に基づいて演算され、 上記第
2の記憶手段に記憶されたビデオ信号を読み出すための読出しァ ドレスを生成す る第 2 の読出しァ ドレス生成手段と
を有することを特徴とする立体映像生成装置。
8. 請求の範囲第 7項の立体映像生成装置において、
上記第 1 の記憶手段に供給される読出しァ ドレス ( XmL , Ymし) は、
XmL = { { - £ P 22 ( £ z P 2 + S 0 ) X
H
十 { J2 y • r 13 - P 2 + r , 2 ( z P z + S o ) Υ
+ ( - r 22 · r , a · Ρ 2 + r 12 · r 23 · Ρ 2 ) 〕
YmL = 〔 I & , - £■ . Ρ ζ Δ χ ) r 23 Ρ
Η
― C r 2 ΐ - r 23 Ρ ζ Δ χ ) · ( Ά ζ Ρ ζ + S ο ) X
+ { - ( £ , - £ ζ Ρ , Δ χ · r , 3 · Ρ ζ
+ ( r , , - r , 3 Ρ ζ Α χ ) · ( 2 Ρ ζ + S D ) Y
+ I r 2 1 - r 23 P 2 Δ x ) · r , 3 P 2
— ( r , ,一 r 13P z 厶 x ) - r 23 · P , } ] ただし、
H = I - r 22 ( £ £ 2 P , A x ) + ( r 2 . - r 23 P ζΔ x ) - £ y) X s
+ I r , 2 ( £ £ 2 P 2 Δ x ) - ( Γ , , - Γ , , Ρ , Δ Χ ) - £ y] Y s
+ I - r , 2 · ( r 2 , - r 2 3 Ρ ζ Δ χ ) + ( r ι ι - r , 3 Ρ ζ Δ χ ) r 22i によって定義され、 上記第 2の記憶手段に供給される読出しア ドレス ( XmR , Y n,R) は、
XmR = X
H
+ { £ y P z + r 12 ( _2 z P z + S 0 ) } Y
+ (一 r • P z + r , 2 · r 23 · P 2 ) :
YmR = 〔 { J2 £ x P , A x ) r 2., P
H
― ( r 2 1 + r 23 P i Δ χ ) - ( £ ζ Ρ , + S η ) ) X
+ { - ( £ , + £ τ Ρ 2 Δ χ · r , 3 · Ρ ,
+ ( r , , + r , 3 Ρ ζ A x ) * ( ^ z P z + S n ) } Y
+ I r 2 , + r 2 3 Ρ , Δ χ ) - r , 3 Ρ ζ
一 ( Γ , , + I" , 3 Ρ Δ χ ) - r , 3 · Ρ , } ] ただし、
Η R = ( - ι- 22 ( £ x + £ t ? Z A x ) + ( r 2 1 + r 23 P z厶 x ) - J2 y) X s + { r 1 2 ( J2 , + je z P zA x ) — ( Γ , , + Ι- Ρ :厶 x ) - S. yl Υ s 十 { - r i a · ( r 2 ) + Γ 23 Ρ ζ Δ χ ) + ( r ι ι + r 1 3 Ρ ζ Δ χ ) r 22 S によって定義され、
X s 及び Y s はラスタスキャンに対応したスク リ ーンア ドレスである ことを特徴とする立体映像生成装置。
9. 請求の範囲第 8項の立体映像生成装置において、
上記第 1 の読出しァ ドレス生成手段及び上記第 1 の読出しァ ドレス生成手段には 、 同じ上記スク リ ーンア ドレス力く、 シーケ ンシャルに供給される
ことを特徴とする立体映像生成装置。
10. 請求の範囲第 7項の立体映像生成装置において、
上記ソ一スビデオ信号を、 上記第 1 の記憶手段及び上記第 2の記憶手段にそれ ぞれ書き込むための書込みァ ドレスを生成する書込みァ ドレス生成手段をさ らに 具え、
上記第 1 の記憶手段及び上記第 2 の記憶手段には、 上記書込みァ ドレス生成手 段から同じ書込みァ ドレスがそれぞれ供給される
ことを特徴とする立体映像生成装置。
1 1. 請求の範囲第 1 項の立体映像生成装置において、
上記左目用プロセッサ及び上記右目用プロセッサには、 上記ソースビデオ信号 に対応したキー信号及び上記ソースビデオ信号の背景に使用されるバックグラウ ン ドビデオ信号がさ らにそれぞれ供給され、
上記左目用プロセッサは、
上記空間的画像変換行列及び上記左目用透視変換行列に基づいて、 上記ソース デォ信号に対して第 1 の空間的画像変換を施すと共に、 上記キー信号に対して 上記第 1 の空間的画像変換と同じ空間的画像変換を施す第 1 の画像変換手段と、 上記第 1 の画像変換手段によって変換されたキー信号に基づいて、 上記第 1 の 画像変換手段によって変換されたビデオ信号と上記バック グラゥ ン ドビデオ信号 とを合成する第 1 のビデオ信号合成手段と
を有し、 上記右目用プロセッサは、
上記空間的画像変換行列及び上記右目用透視変換行列に基づいて、 上記ソ一スビ デォ信号に対して第 2の空間的画像変換を施すと共に、 上記キー信号に対して上 記第 2 の空間的画像変換と同じ空間的画像変換を施す第 2の画像変換手段と、 上記第 2の画像変換手段によって変換されたキ一信号に基づいて、 上記第 2 の 画像変換手段によって変換されたビデオ信号と上記バックグラゥン ドビデオ信号 とを合成する第 2のビデオ信号合成手段と
を有することを特徴とする立体映像生成装置。
12. 請求の範囲第 1項の立体映像生成装置において、 上記左目用プロセッサから出力される出力ビデオ信号を記録媒体に記録すると 共に、 当該記録されたビデオ信号を記録媒体から再生する第 1 の記録再生装置と 上記右目用プロセッサから出力される出力ビデオ信号を記録媒体に記録すると 共に、 当該記録されたビデオ信号を記録媒体から再生する第 2 の記録再生装置と 上記第 1 の記録再生装置から再生された第 1 の再生ビデオ信号からキー信号を 生成する第 1 のキー信号生成手段と、
上記第 2 の記録再生装置から再生された第 2 の再生ビデオ信号からキ一信号を 生成する第 2 のキー信号生成手段と、
をさ らに具え、
上記左目用プロセッサは、
上記第 1 の記録再生装置から再生された第 1 の再生ビデオ信号、 上記第 1 のキ 信号生成手段によって生成されたキー信号及び上記ソースビデオ信号の背景に 使用されるバック グラウ ン ドビデオ信号を受け、
上記空間的画像変換行列及び上記左目用透視変換行列に基づいて、 上記第 1 の 再生ビデオ信号に対して第 1 の空間的画像変換を施すと共に、 上記キー信号に対 して上記第 1 の空間的画像変換と同じ空間的画像変換を施し、
上記第 1 の画像変換手段によって変換されたキー信号に基づいて、 上記第 1 の 画像変換手段によって変換されたビデオ信号と上記バックグラゥ ン ドビデオ信号 とを合成し、
上記右目用プロセッサは、
上記第 2 の記録再生装置から再生された第 2 の再生ビデオ信号、 上記第 2 のキ 一信号生成手段によって生成されたキー信号及び上記バックグラゥ ン ドビデオ信 号を受け、
上記空間的画像変換行列及び上記右目用透視変換行列に基づいて、 上記第 2 の 再生ビデオ信号に対して第 2の空間的画像変換を施すと共に、 上記キー信号に対 して上記第 2 の空間的画像変換と同じ空間的画像変換を施し、 上記第 2の画像変換手段によって変換されたキー信号に基づいて、 上記第 2 の 画像変換手段によって変換されたビデオ信号と上記バックグラゥ ン ドビデオ信号 とを合成する
ことを特徴とする立体映像生成装置。
13. 請求の範囲第 1 2項の立体映像生成装置において、
上記第 1 のキー信号生成手段によって生成されるキー信号と、 上記第 2 のキー 信号生成手段によって生成されるキー信号とは、 同じキー信号である
ことを特徴とする立体映像生成装置。
14. 視覚的に立体効果を得るためのビデオ信号を生成するビデオ信号生成装置に おいて、
ソースビデオ信号によって表される映像を仮想的な空間位置に変換する画像変 換手段と、
.上記画像変換手段によって変換された変換ビデオ信号の映像がある仮想的な空 間位置と、 左目の空間位置とに基づいて、 スク リ ーン面上に透視される映像を表 す左目用ビデオ信号を生成する左目用ビデオ信号生成手段と、
上記変換ビデオ信号の映像がある仮想的な空間位置と、 右目の空間位置とに基 づいて、 スク リ ーン面上に透視される映像を表す右目用ビデオ信号を生成する右 目用ビデオ信号生成手段と
を具えることを特徴とするビデオ信号生成装置。
15. 請求の範囲第 1 4項のビデオ信号生成装置において、
上記画像変換手段は空間的画像変換行列に基づいて上記ソースビデオ信号を変 換し、
上記左目用ビデオ信号生成手段は左目用透視変換行列に基づいて上記左目用ビ デォ信号を生成し、
上記右目用ビデオ信号生成手段は右目用透視変換行列に基づいて上記右目用ビ デォ信号を生成する ことを特徴とするビデオ信号生成装置。
16. 請求の範囲第 1 4項のビデオ信号生成装置において、
ラ スタスキャンに対応したシーケ ンシャルなスク リ ーンァ ドレスを発生させる ためのスク リ ーンァ ドレス発生手段をさ らに具え、
上記左目用ビデオ信号生成手段及び上記右目用ビデオ信号生成手段は上記スク リ ー ンァ ドレスに応答して同時に動作する
ことを特徴とするビデオ信号生成装置。
17. 請求の範囲第 1 5項のビデオ信号生成装置において、
上記空間的画像変換行列 T D は、
1" r r 0
r Γ Γ 0
τ (,
r Γ Γ 0
Figure imgf000050_0001
.よって定義され、
上記左目用透視変換行列 P は.
1 0 0 0
0 1 0 0
P
P 2 厶 X 0 0 P
0 0 0 1 よって定義され、
上記右目用透視変換行列 P R は、 1 0 0 0
0 1 0 0
P R
P 2 Δ x 0 0 P
0 0 0 1 によって定義され、
ただし、 r M ~ r 3 3は空間的な回転変換を表すパラメ ータであり
S. » は X軸方向の平行移動を表すパラメ一夕であり、
& , は y幸由方向の平行移動を表すパラメ一夕であり、
& z は z軸方向の平行移動を表すパラメ 一夕であり、
S n は 2次元的な拡大縮小を表すパラメータであり、
P z は予め設定されたパースペクティブ値であり、
. Δ Xは右目と左目の距離の半分の距離を示す値である
ことを特徴とする立体映像生成装置。
18. 請求の範囲第 1 7項のビデオ信号生成装置において、
上記左目用透視変換行列 P は、
P L = L P„ L によって求められ、 上記右目用透視変換行列 P R は、
P R = L L P 0 によって求められ、
ただし、 行列 Lは、 1 0 0
0 0 0
L =
0 1 0
Δ x 0 1 によって定義され、 行列 L—1は.
1 0 0 0
0 1 0 0
L
0 0 1 0
厶 X 0 0 1 によって定義され、 行列 Ρ。 は.
1 0 0 0
0 1 0 0
Ρ
0 0 1 Ρ
0 0 0 1
によって定義される
ことを特徴とする ビデオ信号生成装置。
19. 請求の範囲第 1 8項のビデオ信号生成装置において、
上記左目用ビデオ信号生成手段は、
上記ソ一スビデオ信号を一時的に記憶する第 1 の記憶手段と、
上記空間的画像変換行列及び上記左目用透視変換行列に基づいて演算され、 上 記第 1 の記憶手段に記憶されたビデオ信号を読み出すための読出しァ ドレスを生 成する第 1 の読出しァ ドレス生成手段と を有し、 上記右目用ビデオ信号生成手段は、
上記ソースビデオ信号を一時的に記憶する第 2の記憶手段と、
上記空間的変換行列及び上記右目用透視変換行列に基づいて演算され、 上記第
2の記憶手段に記憶されたビデオ信号を読み出すための読出しア ドレスを生成す るための第 2の読出しァ ドレス生成手段と
を有することを特徴とする ビデオ信号生成装置。
20. 請求の範囲第 1 9項のビデオ信号生成装置において、
上記第 1 の記憶手段に供袷される読出しア ドレス (XmL、 YmL) は、
X [ ( - J2 - r 23 - P z P 2 + S„ ) X
H
+ { ^ y · Γ , 3 · Ρ 2 + Γ Ι 2 ( ^ ζ Ρ ζ + 8 Ο ) } Υ
+ ( - Γ 22 · Γ 13 · Ρ 2 + Γ Ι 2 · Γ 23 · Ρ Ζ )
YmL = 〔 { Ά £ ζ Ρ ζ Δ χ ) r 23 Ρ
Η
一 ( r 2 , - r P Z A X ) - ( J2 , P 2 + S 0 ) } X
+ ΐ一 ( £ χ P z Δ x - r , 3 · P 2
+ ( r , , - r P Δ x ) · ( JZ z P z + S o ) 1 Y
+ I r 21 - r P z Δ x ) · r ! 3 P ,
一 ( Γ 〗 >一 Γ P Δ x ) · r 23 · P z } ) ただし、
H L = {— r 22 ( ^ i— ·δ ,Ρ :Α χ ) 十 C r 2 , - r 23 P Δ x ) X s
( r , 2 ( J2 x— ^ , P z厶 x ) - ( r , , - r .3 P 2 Δ x ) - Ά y) Υ s + ' - r , 2 · ( r 2i - r 23 P zA x ) + ( r u— r 13P z厶 x ) r 22i によって定義され、 上記第 2の記憶手段に供給される読出しア ドレス (XmR、 Y mR) は、 X [ { - J2 x • r 23 - P i + r 22 ( ^ , P 2 + S o ) X
H
+ { J2 y • r - P i + r 1 2 ( J2 z P i + S o ) } Y
+ ( - r · r ] · P z + r · r 23 · P z ) 〕
Y 〔 { S. , + S, 2 P , Δ x ) r P
H
一 ( r + r P A x ) * ( J2 z P z + S 0 ) i X
+ { - ( Ά x + £ P 2 Δ x · r 13 · P ,
+ ( r , , + r , P A x ) - ( 2 P , + S o ) i Y
+ ( r , + r Ρ Δ x ) · r 1 3 P 2
一 ( r , , + r , Ρ Δ x ) · r · P , t )
ただし、
Η R = {— r 22 ( j2 x+ j2 , P :A x ) + ( r 21 + r 23 P ,Δ x ) - S. y) X s + ( r i2 ( x+ J2 z P zA x ) - ( r , , + r 1 3 P zA x ) - S, y} Y s + < - r , · ( r 2 i + r 23 P IA x ) + ( Γ , , + Γ ^ Ρ Ζ Δ Χ ) r 221 によって定義され、
ただし、 X 及び Y はスク リ ーンア ドレスである
ことを特徴とする ビデオ信号生成装置。
21. 請求の範囲第 2 0項のビデオ信号生成装置において、
上記第 1 の読出しア ドレス生成手段及び上記第 2 の読出しァ ドレス生成手段に は、 同じ上記スク リ ーンア ドレス力く、 ラス夕スキャンに対応するよ う にシーケ ン シャルに供給される
ことを特徵とするビデオ信号生成装置。
22. 請求の範囲第 1 9項のビデオ信号生成装置において、
上記ソースビデオ信号を上記第 1 の記憶手段及び上記第 2 の記憶手段にそれぞ れ書込むための書き込みァ ドレスを生成する書込みァ ドレス生成手段をさ らに具 え、
上記第 1 の記憶手段に供給される書込みァ ドレス及び上記第 2 の記億手段には 、 上記書き込みァ ドレス生成手段から、 同じ書き込みァ ドレスがそれぞれ供給さ れる
ことを特徴とするビデオ信号生成装置。
23. 請求の範囲第 1 5項のビデオ信号生成装置において、
上記左目用ビデオ信号生成手段及び上記右目用ビデオ信号生成手段には、 上記 ソースビデオ信号に対応したキー信号及び上記ソースビデォ信号の背景に使用さ れるバック グラウ ン ドビデオ信号がさ らにそれぞれ供給され、
上記左目用ビデオ信号生成手段は、
上記空間的画像変換行列及び上記左目用透視変換行列に基づいて、 上記ソ一ス ビデオ信号に対して第 1 の空間的画像変換を施すと共に、 上記キー信号に対して _t記第 1 の空間的画像変換と同じ空間的画像変換を施す第 1 の画像変換手段と、 上記第 1 の画像変換手段によって変換されたキー信号に基づいて、 上記第 1 の 画像変換手段によって変換されたビデオ信号と上記バックグラゥン ドビデオ信号 とを合成する第 1 のビデオ信号合成手段と
を有し、 上記右目用ビデオ信号生成手段は、
上記空間的画像変換行列及び上記右目用透視変換行列に基づいて、 上記ソース ビデオ信号に対して第 2の空間的画像変換を施すと共に、 上記キー信号に対して 上記第 2 の空間的画像変換と同じ空間的画像変換を施す第 2の画像変換手段と、 上記第 2の画像変換手段によって変換されたキー信号に基づいて、 上記第 2 の 画像変換手段によって変換されたビデオ信号と上記バックグラゥ ン ドビデオ信号 とを合成する第 2のビデオ信号合成手段と
を有することを特徴とする ビデオ信号生成装置。
24. 請求の範囲第 1 5項のビデオ信号生成装置において、
上記左目用ビデオ信号生成手段から出力される出力ビデオ信号を記録媒体に記 録すると共に、 記録されたビデオ信号を記録媒体から再生する第 1 の記録再生装 置と、
上記右目用ビデオ信号生成手段から出力される出力ビデオ信号を記録媒体に記 録すると共に、 記録されたビデオ信号を記録媒体から再生する第 2の記録再生装 置と、
上記第 1 の記録再生装置から再生された第 1 の再生ビデオ信号からキー信号を 生成する第 1 のキー信号生成手段と、
上記第 2の記録再生装置から再生された第 2の再生ビデオ信号からキー信号を 生成する第 2のキー信号生成手段と
をさ らに具え、
上記左目用ビデオ信号生成手段は、
上記第 1 の記録再生装置から再生された第 1 の再生ビデオ信号、 上記第 1 のキ 一信号生成手段によって生成されたキー信号及び上記ソースビデオ信号の背景に 用されるバック グラウン ドビデオ信号を受け、
上記空間的画像変換行列及び上記左目用透視変換行列に基づいて、 上記第 1 の 再生ビデオ信号に対して第 1 の空間的画像変換を施すと共に、 上記キー信号に対 して上記第 1 の空間的画像変換と同じ空間的画像変換を施し、
上記第 1 の画像変換手段によって変換されたキー信号に基づいて、 上記第 1 の 画像変換手段によって変換されたビデオ信号と上記バックグラゥ ン ドビデオ信号 とを合成し、
上記右目用プロセッサは、
上記第 2の記録再生装置から再生された第 2の再生ビデオ信号、 上記第 2のキ —信号生成手段によって生成されたキー信号及び上記バックグラウン ドビデオ信 号を受け、
上記空間的画像変換行列及び上記右目用透視変換行列に基づいて、 上記第 2の 再生ビデオ信号に対して第 2の空間的画像変換を施すと共に、 上記キー信号に対 して上記第 2の空間的画像変換と同じ空間的画像変換を施し、 上記第 2の画像変換手段によって変換されたキ一信号に基づいて、 上記第 2 の 画像変換手段によって変換されたビデオ信号と上記バックグラゥ ン ドビデオ信号 とを合成する
ことを特徴とするビデオ信号生成装置。
25. 請求の範囲第 2 4項のビデオ信号生成装置において、
上記第 1 のキー信号生成手段によって生成されるキー信号と、 上記第 2 のキー 信号生成手段によって生成されるキー信号は、 同じキー信号である
ことを特徴とするビデオ信号生成装置。
26. 視覚的に立体効果を得るためのビデオ信号を生成するビデオ信号生成方法に おいて、
( a ) ソース ビデオ信号を仮想的な空間位置に変換する第 1 のステップと、
( b ) 上記第 1 のステップにおいて変換された変換ビデオ信号の仮想的な空間位 置及び左目の空間位置に基づいて、 スク リーン面上に透視される左目用ビデオ信 号を生成する第 2 のステップと、
( c ) 上記第 1 のステップにおいて変換された変換ビデオ信号の仮想的な空間位 置及び右目の空間位置に基づいて、 ス ク リ ーン面上に透視される右目用ビデオ信 号を生成する第 3 のステップと
を具えるこ とを特徴とするビデオ信号生成方法。
27. 請求の範囲第 2 6項のビデオ信号生成方法において、
上記第 1 のステツプでは、 オペレー夕の所望の空間的画像変換操作に対応した パラメ一タを有する空間的画像変換行列に基づいて上記ソ一スビデオ信号を変換 し、
上記第 2のステップでは、 予めオペレー夕によって設定されたパラメ ータを有 する左目用透視変換行列に基づいて上記左目用ビデオ信号を生成し、
上記第 3のステップでは、 予めオペレータによって設定されたパラ メ ータを有 する右目用透視変換行列に基づいて上記右目用ビデオ信号を生成する
ことを特徴とするビデオ信号生成方法。
28. 請求の範囲第 2 6項のビデオ信号生成方法において、
ラスタスキャンに対応するシーケンシャルなスク リ ーンァ ドレスを発生させる ための第 4のステップをさ らに有し、
上記第 1 のステップ、 上記第 2のステップ及び上記第 3のステップにおける処 理は、 上記スク リーンァ ドレスに応答して行われる
ことを特徴とする ビデオ信号生成方法。
29. 請求の範囲第 2 7項のビデオ信号生成方法において、
上記空間画像的変換行列 T。 は、
Figure imgf000058_0001
よって定義され、
上記左目用透視変換行列 P L は
1 0 0 0
0 1 0 0
P
P 2 Δ X 0 0 P
0 0 0 1 によって定義され、
上記右目用透視変換行列 P R は、 1 0 0 0
0 1 0 0
P R =
P z 厶 x 0 0 P
0 0 0 1 によって定義され、
ただし、 l- , , "^ !- ^は、 空間的な回転変換を表すパラメ 夕であり
£ X は X軸方向の平行移動を表すパラメ一夕であり、
は y軸方向の平行移動を表すパラメータであり、
ί ζ は ζ軸方向の平行移動を表すパラメ一夕であり、
S 0 は 2次元的な拡大縮小を表すパラメ ータであり、
Ρ , は予め設定されたパースペクティブ値であり、
- Δ Xは右目と左目の距離の半分の距離を示す値である
ことを特徴とする ビデオ信号生成方法。
30. 請求の範囲第 2 9項のビデオ信号生成方法において、
上記左目用透視変換行列 Ρ は、
P L = L Ρ„ L - 1 によって求められ、 上記右目用透視変換行列 P R は.
P R = L - 1 L P 0 によって求められ、
ただし、 行列しは、 1 0 0 0
0 1 0 0
L =
0 0 1 0
Δ x 0 0 1 によって定義され、 行列 L— 1は、
1 0 0
0 0 0
L
0 1 0
Δ χ 0 1 によって定義され、 行列 Ρ η は.
1 0 0 0
0 1 0 0
Ρ ο
0 0 1 Ρ
0 0 0 1 によって定義される
ことを特徴とする ビデオ信号生成方法。
31. 請求の範囲第 3 0項のビデオ信号生成方法において、
上記ソ一スビデオ信号を第 1 の記憶手段及び第 2の記憶手段に一時的にそれぞ れ記憶する第 5のステップと、
上記空間的画像変換行列及び上記左目用透視変換行列に基づいて、 上記第 1 の 記憶手段に記憶されたビデオ信号を読み出すための左目用読出しア ドレスを生成 する第 6のステップと、
上記空間的画像変換行列及び上記右目用透視変換行列に基づいて、 上記第 2の 記憶手段に記憶されたビデオ信号を読み出すための右目用読出しァ ドレスを生成 する第 7のステップと
をさ らに具えることを特徴とする ビデオ信号生成方法。
32. 請求の範囲第 3 1項のビデオ信号生成方法において、
上記左目用の読出しア ド レス (Xmし、 YmL) は、
XmL = ί {- £ r 2 P, + r 22 ( £ z P 2 + S。 ) X
H
· r , 3 · P z + r , 2 ( 2 P 2 + S„ ) I Y s + ( - r 2 2 · r i a · P z + r 1 2 · r 23 · P 2 ) )
YmL = 〔 { £ P , Δ x ) r 23 P
H
一 ( r 2 , - r 23 P 2 A x) · ( £ z P 2 + S 0 ) 1 X
+ ( - ( £ x — P 2 Δ x - r 13 · P z
+ ( r , , - r , 3 P 2 Δ x ) · ( J22 P z + S。 ) } Y
+ 1 r 2 1 - r 2 a P 2 Δ x ) · r , 3 P 2
( r , , - r , 3 P z A x) · r 23 · P 2 \ 〕 ただし、
HL = {— r 22 ( JZx— ·β 2Ρ :Δ χ ) + ( ι- 2,— r 23P :厶 x ) · J2 y ) X + ( r 12 ( J2 XZP Z厶 x ) — ( r i l— ι- 13Ρ2Δ χ) - £ y ] Y + { - r 12 · ( r 2i- r 23 P z A x ) + ( r i , - r , a P z Δ x ) r 2 上記右目用の読出しア ド レス (XmR、 YmR) は、 XmR = ί { - S> x · Γ 23 · Ρ Ζ + Γ 22 ( ^ 2 Ρ , + 8 ο ) X
H
+ { £ Ρ ζ + r , 2 ( Α ζ P z + S a ) Y
+ (一 22 • P , + r 12 · r · P 2 ) )
Y 〔 I S, x + & P 2 Δ x ) r 23 P
H R
一 ( r 21 + r 23 P 2 Δ x ) · ( £ z P z + S o ) X
+ I— ( £ » + SL τ P , Δ x · r 13 · P z
+ ( r , , + r , 3 P z Δ x ) · ( S. t P 2 + S o ) Y
+ { Γ 2 1 + Γ 23 Ρ ζ Δ x ) - r , 3 P z
― ( r , , + r ! 3 P z Δ x ) - r 23 · P 2 1 3
. ただし、
H R = { - r 22 ( J2 x+ J2 2 P z Δ x ) + ( r 21+ Γ 23 Ρ ΖΔ Χ ) · Ά Y) X S + { ι· 12 ( ·β ,+ ·β 2 Ρ ζ厶 x) - ( r , , + r , 3 P ζΔ x ) - S, y} Y s + { - r , 2 · ( r 2i+ r 23P z厶 x) + ( Γ Μ + Γ Ι 3 Ρ 2Δ Χ ) r 22} によって定義され、
ただし、 X S 及び Y s はラス夕 スキャンに対応したシーケ ン シャルに供給され るスク り 一ンァ ドレスである
ことを特徵とするビデオ信号生成方法。
33. 視覚的に立体効果を得るためのビデオ信号を生成するビデオ信号生成方法に おいて、
( a ) ソースビデオ信号、 上記ソースビデオ信号に対応したソースキー信号、 及 び上記ソ一スビデオ信号の背景に使用されるバック グラゥン ドビデオ信号を受け 取る第 1 のステップと、
( b ) 上記ソースビデオ信号及び上記ソースキー信号を仮想的な空間位置に変換 する第 2 のステップと、
( c ) 上記第 2のステップにおいて変換された変換ビデオ信号の仮想的な空間位 置及び変換キー信号の仮想的な空間位置と、 左目の空間位置とに基づいて、 スク リ一ン面上に透視される左目用ビデオ信号及び左目用キー信号をそれぞれ生成す る第 3 のステップと、
( d ) 上記左目用キー信号に基づいて、 上記左目用ビデオ信号及び上記バックグ ラ ウ ン ドビデオ信号を合成する第 4 のステップと、
( e ) 上記第 2のステップにおいて変換された変換ビデオ信号の仮想的な空間位 置及び変換キー信号の仮想的な空間位置と、 右目の空間位置とに基づいて、 ス ク リ ー ン面上に透視される右目用ビデオ信号及び右目用キー信号をそれぞれ生成す る第 5のステップと、
.( f ) 上記右目用キ一信号に基づいて、 上記右目用ビデオ信号及び上記バック グ ラウ ン ドビデオ信号を合成する第 6のステップと
を具えることを特徴とするビデオ信号生成方法。
34. 視覚的に立体効果を得るためのビデオ信号を生成する ビデオ信号生成システ 厶において、
( a ) ソースビデオ信号仮想的な空間位置に変換する第 1 のステップと、
( b ) 上記第 1 のステップにおいて変換された変換ビデオ信号の仮想的な空間位 置及び左目の空間位置に基づいて、 スク リーン面上に透視される左目用ビデオ信 号を生成する第 2 のステップと、
( c ) 上記第 1 のステップにおいて変換された変換ビデオ信号の仮想的な空間位 置及び右目の空間位置に基づいて、 スク リー ン面上に透視される右目用ビデオ信 号を生成する第 3 のステップと、
( d ) 上記左目用ビデオ信号及び上記右目用ビデオ信号を同時にスク リ一ン面上 に表示する第 4 のステップと を具えることを特徴とするビデオ信号生成システム
PCT/JP1996/003027 1995-10-19 1996-10-18 Procede et dispositif de formation d'images en trois dimensions WO1997015150A1 (fr)

Priority Applications (2)

Application Number Priority Date Filing Date Title
GB9709228A GB2309847B (en) 1995-10-19 1996-10-18 Stereoscopic image generation method and apparatus thereof
US08/860,270 US5959663A (en) 1995-10-19 1996-10-18 Stereoscopic image generation method and apparatus thereof

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP29737195 1995-10-19
JP7/297371 1995-10-19

Publications (1)

Publication Number Publication Date
WO1997015150A1 true WO1997015150A1 (fr) 1997-04-24

Family

ID=17845630

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP1996/003027 WO1997015150A1 (fr) 1995-10-19 1996-10-18 Procede et dispositif de formation d'images en trois dimensions

Country Status (3)

Country Link
US (2) US5959663A (ja)
GB (1) GB2309847B (ja)
WO (1) WO1997015150A1 (ja)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014506039A (ja) * 2010-12-06 2014-03-06 テラネックス システムズ インコーポレイテッド 疑似3d強化遠近法を用いた方法及び装置
US20140375663A1 (en) * 2013-06-24 2014-12-25 Alexander Pfaffe Interleaved tiled rendering of stereoscopic scenes
WO2019218887A1 (zh) * 2018-04-27 2019-11-21 苏州新光维医疗科技有限公司 二维图像转换成三维图像的方法、装置及三维成像系统
ES2813426A1 (es) * 2018-05-23 2021-03-23 Scivita Med Tech Co Ltd Metodo y dispositivo de procesamiento de imagenes y sistema de formacion de imagenes tridimensionales

Families Citing this family (50)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1997015150A1 (fr) * 1995-10-19 1997-04-24 Sony Corporation Procede et dispositif de formation d'images en trois dimensions
JP3519594B2 (ja) * 1998-03-03 2004-04-19 Kddi株式会社 ステレオ動画像用符号化装置
WO2000019265A1 (de) * 1998-09-30 2000-04-06 Siemens Aktiengesellschaft Anordnung und verfahren zur stereoskopischen darstellung eines objekts
NO308925B1 (no) * 1999-03-15 2000-11-13 Dimension Technologies As FremgangsmÕte og apparat for stereoprojeksjon av bilder
USD445818S1 (en) 1999-05-20 2001-07-31 Point Grey Research Inc. Trinocular stereo vision camera
KR100381817B1 (ko) * 1999-11-17 2003-04-26 한국과학기술원 제트버퍼를 이용한 입체영상 생성방법 및 기록매체
GB2377094B (en) * 2001-06-29 2003-07-30 Paul Morris Power converter
JP3945279B2 (ja) * 2002-03-15 2007-07-18 ソニー株式会社 障害物認識装置、障害物認識方法、及び障害物認識プログラム並びに移動型ロボット装置
US7804995B2 (en) 2002-07-02 2010-09-28 Reald Inc. Stereoscopic format converter
DE10242262A1 (de) * 2002-09-12 2004-03-25 Daimlerchrysler Ag Stereo-Nachtsichtsystem für Fahrzeuge
US7643025B2 (en) * 2003-09-30 2010-01-05 Eric Belk Lange Method and apparatus for applying stereoscopic imagery to three-dimensionally defined substrates
US7477140B1 (en) * 2003-12-26 2009-01-13 Booth Kenneth C See-through lighted information display
US7643672B2 (en) * 2004-10-21 2010-01-05 Kazunari Era Image processing apparatus, image pickup device and program therefor
KR100932977B1 (ko) * 2005-07-05 2009-12-21 삼성모바일디스플레이주식회사 입체 영상 표시 장치
KR100913173B1 (ko) * 2005-07-05 2009-08-19 삼성모바일디스플레이주식회사 3d 그래픽 처리장치 및 이를 이용한 입체영상 표시장치
US8279221B2 (en) * 2005-08-05 2012-10-02 Samsung Display Co., Ltd. 3D graphics processor and autostereoscopic display device using the same
EP1943558A4 (en) * 2005-09-16 2010-06-09 Stereographics Corp STEREOSCOPIC FORMAT CONVERTER
US8130227B2 (en) * 2006-05-12 2012-03-06 Nvidia Corporation Distributed antialiasing in a multiprocessor graphics system
US7670004B2 (en) * 2006-10-18 2010-03-02 Real D Dual ZScreen® projection
KR101339870B1 (ko) * 2007-07-26 2013-12-11 삼성전자주식회사 영상처리장치 및 영상처리방법
USD624952S1 (en) 2008-10-20 2010-10-05 X6D Ltd. 3D glasses
USD603445S1 (en) 2009-03-13 2009-11-03 X6D Limited 3D glasses
USD666663S1 (en) 2008-10-20 2012-09-04 X6D Limited 3D glasses
USRE45394E1 (en) 2008-10-20 2015-03-03 X6D Limited 3D glasses
CA2684513A1 (en) 2008-11-17 2010-05-17 X6D Limited Improved performance 3d glasses
US8542326B2 (en) 2008-11-17 2013-09-24 X6D Limited 3D shutter glasses for use with LCD displays
USD646451S1 (en) 2009-03-30 2011-10-04 X6D Limited Cart for 3D glasses
JP4915459B2 (ja) * 2009-04-03 2012-04-11 ソニー株式会社 情報処理装置、情報処理方法、及び、プログラム
US20100265315A1 (en) * 2009-04-21 2010-10-21 Panasonic Corporation Three-dimensional image combining apparatus
USD672804S1 (en) 2009-05-13 2012-12-18 X6D Limited 3D glasses
USD650956S1 (en) 2009-05-13 2011-12-20 X6D Limited Cart for 3D glasses
USD669522S1 (en) 2010-08-27 2012-10-23 X6D Limited 3D glasses
USD671590S1 (en) 2010-09-10 2012-11-27 X6D Limited 3D glasses
USD692941S1 (en) 2009-11-16 2013-11-05 X6D Limited 3D glasses
EP2355526A3 (en) 2010-01-14 2012-10-31 Nintendo Co., Ltd. Computer-readable storage medium having stored therein display control program, display control apparatus, display control system, and display control method
JP5898842B2 (ja) 2010-01-14 2016-04-06 任天堂株式会社 携帯型情報処理装置、携帯型ゲーム装置
USD662965S1 (en) 2010-02-04 2012-07-03 X6D Limited 3D glasses
JP5800501B2 (ja) 2010-03-12 2015-10-28 任天堂株式会社 表示制御プログラム、表示制御装置、表示制御システム、及び、表示制御方法
US8633947B2 (en) 2010-06-02 2014-01-21 Nintendo Co., Ltd. Computer-readable storage medium having stored therein information processing program, information processing apparatus, information processing system, and information processing method
US8384770B2 (en) 2010-06-02 2013-02-26 Nintendo Co., Ltd. Image display system, image display apparatus, and image display method
EP2395768B1 (en) 2010-06-11 2015-02-25 Nintendo Co., Ltd. Image display program, image display system, and image display method
USD664183S1 (en) 2010-08-27 2012-07-24 X6D Limited 3D glasses
JP5739674B2 (ja) 2010-09-27 2015-06-24 任天堂株式会社 情報処理プログラム、情報処理装置、情報処理システム、および、情報処理方法
US8854356B2 (en) 2010-09-28 2014-10-07 Nintendo Co., Ltd. Storage medium having stored therein image processing program, image processing apparatus, image processing system, and image processing method
KR101750047B1 (ko) * 2010-10-11 2017-06-22 삼성전자주식회사 3차원 영상 제공 및 처리 방법과 3차원 영상 제공 및 처리 장치
US8913664B2 (en) * 2011-09-16 2014-12-16 Sony Computer Entertainment Inc. Three-dimensional motion mapping for cloud gaming
KR101890622B1 (ko) 2011-11-22 2018-08-22 엘지전자 주식회사 입체영상 처리 장치 및 입체영상 처리 장치의 칼리브레이션 방법
JPWO2013099290A1 (ja) * 2011-12-28 2015-04-30 パナソニック株式会社 映像再生装置、映像再生方法、映像再生プログラム、映像送信装置、映像送信方法及び映像送信プログラム
USD711959S1 (en) 2012-08-10 2014-08-26 X6D Limited Glasses for amblyopia treatment
KR102581470B1 (ko) * 2017-11-22 2023-09-21 삼성전자주식회사 영상 데이터를 처리하는 방법 및 장치

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH07182533A (ja) * 1993-12-22 1995-07-21 Sanyo Electric Co Ltd 2次元画像の3次元化方法
JPH07200870A (ja) * 1993-12-28 1995-08-04 Sharp Corp 立体視用3次元画像生成装置
JPH07226958A (ja) * 1994-02-16 1995-08-22 Sanyo Electric Co Ltd 立体映像表示方式

Family Cites Families (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4734756A (en) * 1981-12-31 1988-03-29 3-D Video Corporation Stereoscopic television system
FR2619664B1 (fr) * 1987-08-20 1990-01-05 Allio Pierre Procede et installation pour la realisation d'images en relief
NL8800595A (nl) * 1988-03-10 1989-10-02 Philips Nv Weergeef- en opneeminrichting voor stereoscopische beeldweergave.
US4943852A (en) * 1988-10-06 1990-07-24 Eclectica, Inc. Stereoscopic converter assembly for closed circuit 2-D television system
US5130794A (en) * 1990-03-29 1992-07-14 Ritchey Kurtis J Panoramic display system
US5193000A (en) * 1991-08-28 1993-03-09 Stereographics Corporation Multiplexing technique for stereoscopic video system
GB2267007B (en) * 1992-04-24 1995-09-13 Sony Broadcast & Communication Video special effect apparatus and method
DE4228111C1 (ja) * 1992-08-25 1993-06-17 Ice Oelsnitz Gmbh, O-9920 Oelsnitz, De
US5495576A (en) * 1993-01-11 1996-02-27 Ritchey; Kurtis J. Panoramic image based virtual reality/telepresence audio-visual system and method
US5469576A (en) * 1993-03-22 1995-11-21 International Business Machines Corporation Front end for file access controller
DE69417824D1 (de) * 1993-08-26 1999-05-20 Matsushita Electric Ind Co Ltd Stereoskopischer Abtastapparat
US5510832A (en) * 1993-12-01 1996-04-23 Medi-Vision Technologies, Inc. Synthesized stereoscopic imaging system and method
JPH0879800A (ja) * 1994-09-06 1996-03-22 Olympus Optical Co Ltd 映像表示装置
US5905525A (en) * 1995-07-13 1999-05-18 Minolta Co., Ltd. Image display apparatus having a display controlled by user's head movement
WO1997015150A1 (fr) * 1995-10-19 1997-04-24 Sony Corporation Procede et dispositif de formation d'images en trois dimensions
WO1997023097A2 (en) * 1995-12-19 1997-06-26 Philips Electronics N.V. Parallactic depth-dependent pixel shifts

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH07182533A (ja) * 1993-12-22 1995-07-21 Sanyo Electric Co Ltd 2次元画像の3次元化方法
JPH07200870A (ja) * 1993-12-28 1995-08-04 Sharp Corp 立体視用3次元画像生成装置
JPH07226958A (ja) * 1994-02-16 1995-08-22 Sanyo Electric Co Ltd 立体映像表示方式

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014506039A (ja) * 2010-12-06 2014-03-06 テラネックス システムズ インコーポレイテッド 疑似3d強化遠近法を用いた方法及び装置
US20140375663A1 (en) * 2013-06-24 2014-12-25 Alexander Pfaffe Interleaved tiled rendering of stereoscopic scenes
WO2019218887A1 (zh) * 2018-04-27 2019-11-21 苏州新光维医疗科技有限公司 二维图像转换成三维图像的方法、装置及三维成像系统
ES2813425A1 (es) * 2018-04-27 2021-03-23 Scivita Med Tech Co Ltd Metodo y dispositivo para convertir una imagen 2d en una imagen 3d y sistema de formacion de imagenes 3d
US11589030B2 (en) 2018-04-27 2023-02-21 Scivita Medical Technology Co., Ltd. Method and device for converting 2D image into 3D image and 3D imaging system
ES2813426A1 (es) * 2018-05-23 2021-03-23 Scivita Med Tech Co Ltd Metodo y dispositivo de procesamiento de imagenes y sistema de formacion de imagenes tridimensionales

Also Published As

Publication number Publication date
GB9709228D0 (en) 1997-06-25
GB2309847B (en) 1999-09-15
US5959663A (en) 1999-09-28
GB2309847A (en) 1997-08-06
US6160574A (en) 2000-12-12

Similar Documents

Publication Publication Date Title
WO1997015150A1 (fr) Procede et dispositif de formation d&#39;images en trois dimensions
US6747610B1 (en) Stereoscopic image display apparatus capable of selectively displaying desired stereoscopic image
US5835133A (en) Optical system for single camera stereo video
US7295699B2 (en) Image processing system, program, information storage medium, and image processing method
US20080246757A1 (en) 3D Image Generation and Display System
JP2000215311A (ja) 仮想視点画像生成方法およびその装置
JPH05143709A (ja) ビデオ効果装置
CA2302199A1 (en) Image processing method and apparatus
JP2012216979A (ja) 画像処理装置、画像処理方法及び画像処理プログラム
WO2022227753A1 (zh) 用于双目畸变校正的显示系统、显示方法及车载系统
US20040212612A1 (en) Method and apparatus for converting two-dimensional images into three-dimensional images
JP3126575B2 (ja) 立体視用3次元画像生成装置
JP2000156876A (ja) 3次元表示方法及び装置
WO2000038434A1 (en) Method and apparatus for converting monoscopic images into stereoscopic images
JP2009077276A (ja) 画像生成方法、立体視用印刷物、立体視用印刷物の製造方法及びプログラム
JP2004178579A (ja) 立体視用印刷物の製造方法、立体視用印刷物
JP2709066B2 (ja) カメラ撮影画像と計算機生成画像の合成方法
KR20080034419A (ko) 3d 영상 생성 및 디스플레이 시스템
JP2008167310A (ja) 裸眼立体視画像処理方法および装置ならびに演算プログラムを記録した記録媒体
JP2004207780A (ja) 写真撮影装置
JPH10336703A (ja) 画像合成装置、方法および記憶媒体
JP2908799B2 (ja) 立体視用画像作成方法および装置
JP2000030080A (ja) Vrシステム
JP2000067265A (ja) 画像処理装置およびその方法
JPH09116932A (ja) 特殊効果装置

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): GB JP US

WWE Wipo information: entry into national phase

Ref document number: 08860270

Country of ref document: US

点击 这是indexloc提供的php浏览器服务,不要输入任何密码和下载