JP2004104742A - Method and apparatus for generating three-dimensional video data - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a soft method and an apparatus for creating three-dimensional (3D) video rich in the sense of depth simultaneously with video which jumps out forwards. <P>SOLUTION: Video for a left/right eye is alternately displayed on a display screen and an observer recognizes 3D video with eyeglasses which are opened/closed synchronously with the video. In order to create left/right video data, first of all, pixels comprising 2D video data have depth data in addition to coordinate data. A predetermined value of the depth data is set as a basic position P where optical axes of left and right eyes cross each other. An area wherein the depth data are smaller than the basic position is defined as positive areas B1 and B2 and an area where the depth data are greater than the basic position P is defined as negative areas B3 and B4. Forward proportional data L1 and L2 for the left eye are created from the positive areas by a shift device and next, data for the left eye are created from the basic area P. Further, backward inversely proportional data L3 and L4 for the left eye are created from the negative areas by the shift device. Such data are defined as data for the left eye. Data for the right eye are similarly created. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a method and an apparatus for generating stereoscopic image data, and more particularly to a method and an apparatus for generating a stereoscopic image from a video image captured by a television camera or an image generated by computer drawing.
[0002]
[Prior art]
With the recent development of computer drawing and the spread of personal computers and video games, it has become possible to appreciate various surreal images.
In 3D, the crossing method is often used, but by crossing, a three-dimensional effect is obtained by approaching the image viewed by the human eye. However, when judging by observing a back object and a front object at the same time for a game, in the conventional method, the viewpoint must be adjusted to one of the objects. Distortion occurred and symptoms such as 3D sickness appeared.
Even in a two-dimensional image, it is possible to make a seemingly three-dimensional image appear by moving the viewpoint or rotating the target image by using a lot of shading and perspective expressions. Furthermore, it is possible to project a three-dimensional stereoscopic image by separating the images that enter the left and right eyes as left and right data, displaying them alternately on the display screen, and viewing these independently with the left and right eyes. Was.
[0003]
[Problems to be solved by the invention]
The three-dimensional stereoscopic video mainly generates a depth with respect to the display screen, and the video that jumps out to the viewer on the display screen is not supplied so much.
The conventional method and apparatus for creating a video projecting to the front have insufficient functions as a software method, and are insufficient for creating a stereoscopic video that appeals strongly to an observer in a wide variety of video ranges.
SUMMARY OF THE INVENTION It is an object of the present invention to provide a software method and apparatus that have a sufficient function of creating an image with a rich sense of depth at the same time as an image that protrudes toward the user.
The new technology is designed so that the object in front and the object in the back can be focused even when viewed at the same time or separately from each other. Can be expressed very far at the same time.
[0004]
[Means for Solving the Problems]
The present invention has been made to solve the above-described problem. In the first invention, a left-eye image and a right-eye image corresponding to the left and right eyes are alternately displayed on an electronic display screen. A method of generating stereoscopic video data provided to a stereoscopic video display device in which an observer recognizes a stereoscopic video by viewing the video by putting glasses on the left and right eyes to open and close in synchronization with the display of Each pixel constituting the video data has depth data whose numerical value changes from the near view to the distant view in addition to the coordinate data, and a predetermined value of the depth data is used as a basic position where the optical axes of the left and right eyes intersect. The central processing unit sets an area in which the depth data is smaller than the basic position as a positive area, sets the basic position in which the depth data is constant as a basic area, and sets the depth data larger than the basic position. The positive region is read out of the divided memory, and the positive direction proportional data for the left or right eye is created by a shift device. By reading the basic area to create left or right eye data, further reading the negative area to create left or right eye reverse inverse proportional data with the shift device, these left or right eye data Storing in the memory for the left or right eye, reading the positive area again from the divided memory to create the right or left eye positive proportional data with the shift device, and then reading the basic area and reading the right area for the right or left eye Create data, further read the negative area, create reverse inverse proportional data for the right or left eye with the shift device, these right or left eye data to the right or left And storing, each memory for eyes.
[0005]
In the second invention, the positive-direction proportional data for the left or right eye of the positive area is such that the smaller the depth data, the larger the amount of shift of the coordinate data to the right or left, and the larger the depth data, the more the right-hand proportional data of the coordinate data. Or, the amount of shift to the left is made smaller, the data for the left or right eye of the basic area is created without the amount of shift, and the inverse inverse proportional data for the left or right eye of the negative area is smaller as the depth data is smaller. The shift device creates a small amount of displacement of the coordinate data to the left or right, and increases the amount of displacement of the coordinate data to the left or right as the depth data increases,
The right or left eye positive direction proportional data of the positive area is such that the smaller the depth data, the larger the amount of shift of the coordinate data to the left or right, and the larger the depth data, the larger the amount of shift of the coordinate data to the left or right. The data for the right or left eye of the basic area is created without a shift amount, and the inverse inverse proportional data for the right or left eye of the negative area is the right or left of the coordinate data as the depth data is smaller. The shift device is characterized in that the shift device creates the amount of shift to the right or left of the coordinate data as the depth data increases, and the amount of shift to the right or left increases as the depth data increases.
[0006]
In the third invention, an input device for setting a predetermined value of the depth data as a basic position at which the optical axes of the left and right eyes intersect, an area where the depth data is smaller than the basic position as a positive area, and the depth data is fixed. A central processing unit that divides the basic position into three as a basic region, a region where the depth data is larger than the basic position as a negative region, a divided memory that distinguishes and stores these regions, and a command from the central processing unit. By reading the positive area from this divided memory to create left or right eye positive direction proportional data, then read the basic area to create left or right eye data, and further read the negative area Along with creating reverse inverse proportional data for the left or right eye, re-reading the positive area from the divided memory and reconstructing the positive proportional data for the right or left eye. And then read the basic area to create data for the right or left eye, and further read the negative area to create reverse inverse proportional data for the right or left eye, and a shift device for the left or right eye. It is characterized by comprising a left or right eye memory for storing data and a right or left eye memory for storing the right or left eye data.
[0007]
In the fourth invention, the shift device increases the amount of displacement of the coordinate data to the right or left as the depth data is smaller, and the larger the depth data, the more the positive direction proportional data for the left or right eye of the positive region is. The coordinate data is created with a small amount of shift to the right or left, the data for the left or right eye of the basic area is created without a shift, and the inverse inverse proportional data for the left or right eye of the negative area is depth. The smaller the data, the smaller the shift amount of the coordinate data to the left or right, and the larger the depth data, the larger the shift amount of the coordinate data to the left or right. In the positive proportional data, the smaller the depth data, the larger the amount of displacement of the coordinate data to the left or right, and the larger the depth data, the more left or right the coordinate data. The amount of displacement is made smaller, the data for the right or left eye of the basic area is created without the amount of displacement, and the inverse inverse proportional data for the right or left eye of the negative area is the coordinate data as the depth data is smaller. It is characterized in that the amount of displacement to the right or left is reduced and the amount of displacement of the coordinate data to the right or left is increased as the depth data increases.
[0008]
BEST MODE FOR CARRYING OUT THE INVENTION
In humans, even if the same object is present in the left and right eyes, an image is formed with a parallax in which the positions of both eyes are shifted in the horizontal direction. As shown in Fig. 1, when a human is looking at infinity such as the sun, a distant mountain, or a river, the eyeball moves and the optical axes of the left and right eyes become parallel, and what can be seen by the left eye moves to the right. And the same thing seen by the right eye is seen moving to the left. In FIG. 2, when the optical axes are parallel to each other, the parallax S1 of the object near the eye is larger than the parallaxes S2 and S3 of the distant object. S1>S2> S3 where the parallax gradually decreases as the distance increases.
[0009]
However, when looking at an object 100 meters away from the eye, the eyeball moves so as to face that direction in FIG. 3, and the left and right eyes form a convergence angle T with respect to the object being viewed. The convergence angle T1 of the object near the eye is larger than the convergence angle T2 of the distant object, T1> T2. The human brain recognizes the three-dimensional image of the object based on the convergence angle T in FIG. 3 and the parallax S in FIG. 2 and determines perspective.
Now, in FIG. 4, consider a case in which objects B1 and B2, and then a transparent plate P on which characters are drawn and objects B3 and B4 are sequentially arranged so as to gradually move away from the front of the human face F.
[0010]
When the eyes focus on the characters on the transparent plate P, the optical axes of the left and right eyes intersect on the object (transparent plate) P. However, for the objects B1 and B2 in front of the transparent plate P, the left and right parallax What you see with your left eye moves right, and what you see with your right eye looks like it moves left.
FIG. 4 shows a state in which the left eye is thereafter covered and the right eye is seen, and for the objects B1 and B2 in front of the transparent plate P, the right and left eyes are the same as in the case where the optical axes of the eyes in FIG. The same objects B1 and B2 visible to the eyes can be seen moving to the left, respectively, and the displacement L1 of the object B1 near the eye is larger than the displacement L2 of the distant object B2. As the distance increases, the amount of displacement (one element of parallax) gradually decreases (L1> L2, which is defined as direct proportion).
[0011]
In FIG. 4, after that, when the right eye is covered and the left eye looks at these, in the objects B1 and B2 in front of the transparent plate P, the same objects B1 and B2 visible to the left eye are seen moving to the right respectively, and Although not shown, it can be easily estimated that the displacement of the object B1 near the eye is larger than the displacement of the object B2 far from the eye.
The sum of the displacement of the same object B1 as seen by the left eye and the displacement of the same object B1 as seen by the right eye is the parallax. In the following description, this displacement will be described as parallax.
[0012]
Now, in FIG. 4, beyond the intersection (transparent plate P), the same objects B3 and B4 seen by the right eye can be seen to have moved to the right in opposite directions. Further, beyond the intersection, the displacement L3 of the object B3 at a position close to the right eye (transparent plate P) is smaller than the displacement L4 of the distant object B4. The parallax gradually increases as the distance from the eye (intersection / transparent plate P) increases (L3 <L4 is inversely proportional).
Although not shown, the same objects B3 and B4, which are similarly visible to the left eye, can be seen to move to the left, respectively, and the amount of displacement can be easily estimated to be inversely proportional.
[0013]
In other words, the direction of displacement and the amount of displacement are reversed before and after the intersection (transparent plate P), and the convergence angle cannot be defined beyond this point.
This is because when you look at your finger with the PC screen etc. in front and bring your finger close to your eyes from the screen side and focus your eyes on your fingers, the PC screen that does not move will be out of focus, but it will fill the field of view greatly It is understandable because it spreads out and can be seen.
[0014]
Generally, an image drawn by a computer is digital image data. Assuming that this data is displayed on a screen of a personal computer, a pixel Pk shown in FIG.
The pixel Pk is composed of normal two-dimensional position data Xk and Yk on the screen, color data Ck (including gradation), depth / far data Zk, and data W related to surrounding pixels. The related data W indicates the relationship of a continuous picture with the four adjacent pixels in the upper, lower, left, and right directions.
[0015]
The digital image data of one screen can be defined as the sum of pixels Pk (Xk, Yk, Ck, Zk, W).
In the image of the two-dimensional bicycle drawn by the perspective view shown in FIG. 7, the image in which the steering wheel 41 is protruded toward the user, the image of the rear wheel 42 being the rear of the image are displayed at the display screen position (basic position) when the seat part 40 is viewed from the observer. 3D stereoscopic video is generated.
Although the data Zk of the depth perspective can theoretically be set at infinity, it is meaningless on display and time is wasted in the calculation, so it is limited to a predetermined finite value. For example, an area watched by an ordinary human is set as a basic position at a predetermined value, for example, Zk = “0”. The position of an image viewed immediately from the basic position Zk = “0” is positive, and the basic position Zk = “0”. Set the video farther away from the camera to a negative value. A negative value (finite value) at the back is determined in consideration of the balance of the image in accordance with the positive value of the front image, and this is set to infinity.
[0016]
The present invention focuses on the characteristics of the human eye and the digital image data, performs data processing on a two-dimensional image so as to generate the characteristics of the eyes, and generates three-dimensional data. Then, on a stereoscopic video display device, a convergence angle and a parallax are generated electronically in the displayed video so that the observer can recognize the stereoscopic video.
[0017]
Embodiment 1 FIG.
Hereinafter, the stereoscopic image generation method and apparatus of the present invention will be described by taking as an example a case where the apparatus used is a desktop computer (personal computer) and also serves as an image display apparatus that displays a stereoscopic image created by the personal computer.
Although the video display device normally accompanies an audio signal, the audio signal circuit is not directly related to the present invention, and thus the description is omitted.
In FIG. 6, the personal computer (stereoscopic image display device) includes a display device 2, glasses 4, a keyboard 5, a mouse 6, an infrared light emitting device 3, and a logic circuit unit 1.
The logic circuit unit 1 is a main body unit, and includes a video input unit which is a reading device for a CD-ROM (disk-shaped high-density reading storage medium), FD (floppy disk), DVD, VTR, or the like. HD (hard disk).
[0018]
Various kinds of viewing software are stored in the CD-ROM or the like in advance, and the stereoscopic video generation program is stored in the internal HD in advance.
In order to freely use the two-dimensional software pack, it is better to store the program mainly in the internal HD.
The logic circuit section 1 includes a CPU 23, a program memory 9, a left memory 14, a right memory 15, a basic memory 16, a divided memory 17, a copy memory 18, a work memory 26, and a storage memory 29 in FIG.
[0019]
The program memory 9 and the storage memory 29 use an internal HD, and since these are read out less frequently, the program memory 9 previously stores a stereoscopic image generation program that is repeatedly used for each operation.
The left memory 14, the right memory 15, the basic memory 16, the divided memory 17, the copy memory 18 and the work memory 26 are preferably semiconductor memories, and during storage, the stored contents are repeatedly read and written by the CPU 23 at high speed. The storage memory 29 stores the contents of the left memory 14 and the right memory 15.
[0020]
Further, the logic circuit unit 1 includes a changeover switch 7, a digital image generation unit 8, a shift circuit 20, a left shift setting unit 19, a right shift setting unit 20, a distribution switch 25, a readout switch 10, a timing generator 22, and a video supply unit 27. And a digital / analog converter (D / A) 30.
The shift circuit 20, the left shift setting device 19, and the right shift setting device 20 create the above-described displacement amount and left / right position shift (shift). Adjust the volume using setters 19,21.
The digital image generator 8 includes an analog-to-digital converter (A / D), and converts an analog video image into a sum of pixels Pk (Xk, Yk, Ck, Zk, W), which is the digital image data, for each screen. .
[0021]
The read switch 10 switches between the left memory 14 and the right memory 15 in response to a timing signal P from a timing generator 22 that operates according to a command from the CPU 23 and reads the data. The timing signal P is also sent to the infrared light emitting device 3 shown in FIG.
When the changeover switch 7 selects the digital image generator 8 on the VTR side from the image supply unit 27, digital image data is reproduced from a commercially available video tape or the like (normal 2D movie or sports image). When the changeover switch 7 selects a DVD or the like from the video input unit 27, the digital image data is directly read out therefrom.
[0022]
Digital image data of one screen unit (sum of pixels Pk (Xk, Yk, Ck, Zk, W)) is temporarily moved to the basic memory 16. Next, the video data is divided by the CPU 23, divided into predetermined sections in the division memory 17, and stored. Thereafter, the divided video data is transferred to the memory 18 by copying the divided partial data individually.
[0023]
A method of creating stereoscopic video data using the two-dimensional bicycle video shown in FIG. 7 as an example will be described in detail with reference to the flow of FIG. The image of the two-dimensional bicycle shown in FIG. 7 is constituted by digital image data, and includes the pixel Pk (Xk, Yk, Ck, Zk, W) data.
In step S1, the CPU 23 instructs the changeover switch 7 to read the first screen from the image supply unit 27, moves the data to the basic memory 16, and displays the two-dimensional digital image on the display device 2 therefrom.
[0024]
In step S2, the operator who observes the two-dimensional digital image designates the above-mentioned basic position in, for example, the seat portion 40 using the mouse 6 or the keyboard 5 in the image.
Then, in step S3, the CPU 23 normalizes the pixel Pk (Xk, Yk, Ck, Zk, W) data based on the seat section 40 in accordance with the generation program read from the program memory 9, ie, this digital image data In this case, the area near the seat section 40 is set as a Zk = “0” area, the area near the front handle section 41 is set as a “positive” area, and the area near the rear rear wheel section 42 is set as a “negative” area. Of course, even in each area, for example, the “correct” area, the perspective data Zk is not constant and has a perspective.
[0025]
The “positive” region near the handle portion 41 and the “negative” region near the rear wheel portion 42 have a predetermined area and depth data, while the Zk = “0” region near the seat portion 40 has an area. Have no depth data.
One piece of digital image data divided into three is transferred to the division memory 17 with the division code.
In step S4, the video to be created is determined to be right or left (either may be first).
In step S5, the "positive" area near the front handle portion 41 is read out from the divided memory 17 and transferred to the memory 18, and in step S6, for example, the shift amount of the right-direction and direct-right proportional shift to the right is shifted by the shift circuit 20. Performed by
[0026]
In the Zk = “positive” region near the handle portion 41, with respect to the object (video) data at a position closer to the right eye, the amount of displacement L1 to the left is larger than the displacement of a distant object, as in FIG. L1> L2 in which the displacement amount L2 is gradually reduced (in the vicinity of the seat portion 41). Note that the proportional constant of the amount of displacement due to perspective can be changed by the left shift setting unit 19, and is empirically set so that the final three-dimensional effect is best. In step S7, the CPU 23 determines the destination by the distribution switch 25 and stores the created right-eye front data in the right memory 15.
In step S8, for the "positive" area near the handle portion 41, it is checked whether the processing has been completed for the depth of all pixel Pk (Xk, Yk, Ck, Zk, W) data. I do. If the processing has been completed, in step S9, non-change processing is performed on the individual position Zk = "0" area near the seat section 40, and this data is also stored in the right memory 15 in step S10. In the individual position “0” area, the perspective data is constant, but has area data.
[0027]
In step S11, the Zk = “negative” area near the rear rear wheel portion 42 is read out from the divided memory 17 and transferred to the memory 18, and in step S12, the rightward displacement for the right eye and the calculation of the reverse amount inversely proportional are shifted. This is performed by the circuit 20.
In the “negative” region near the rear wheel portion 42, with respect to the object (video) data at a position closer to the right eye, the displacement L3 to the right is smaller than the displacement of the distant object, and the further the distance, the farther (the rear wheel) (Backward) L3 <L4 in which the displacement L4 is gradually increased. Also here, the inverse proportional constant of the amount of displacement due to perspective can be changed by the right shift setting unit 21 and is set empirically so that the final three-dimensional effect is best.
[0028]
In step S13, the created right-eye depth (negative area) data is stored in the right memory 15 via the distribution switch 25. In step S14, for the "negative" area near the rear wheel portion 42, it is checked whether the processing has been completed for the depth of all pixel Pk (Xk, Yk, Ck, Zk) W) data. do.
If the processing has been completed, the process proceeds to step S15, and processing for generating left-eye transition data is performed. This processing is the same as the processing from step S5 to step S14, except that in the corresponding processing, the right is left (and the left is right), the transition direction is reversed, and the data storage position is the left memory 14. Since the processing is performed, the description is omitted.
[0029]
Thus, the left-eye and right-eye data of the first still image are stored in the left memory 14 and the right memory 15 with the respective pages. In step S16, the CPU 23 transfers the data in the left memory 14 and the right memory 15 to the storage memory 29. In step S17, the above steps S2 to S16 are performed for the video of the next page.
By repeating this operation, stereoscopic video data for a required page can be created and stored in the storage memory 29. In the case of a moving image, the number of pages (pages) increases according to the length of the story.
Left and right stereoscopic video data from the storage memory 29 is recorded as digital signals on a CD-ROM or digital magnetic tape, and can be reproduced on another stereoscopic video display device to view a stereoscopic video.
The left and right shift setting units 19 and 21 may be adjusted by providing knobs for parts on the outer surface of the logic circuit unit 1, or may be set by logic on a program.
[0030]
The operation of displaying the created stereoscopic video data on the screen of the display device 2 of the desktop computer (personal computer) in FIG. 5 to view the stereoscopic video will be described.
The screen of the personal computer is a non-interlaced high-resolution normal scan, and the scan speed is set to, for example, 80 Hz.
In accordance with the read program, the CPU 23 reads left and right stereoscopic video data from the storage memory 29 and transfers them to the left and right memories 14 and 15, respectively. Next, the CPU 23 alternately reads, for example, left video data from the left memory 19 in the first 1/80 second, and reads right video data from the right memory 15 in the next 1/80 second, and sends them to the display device 2.
[0031]
The left video data is read out in the third 1/80 second, and the right video data is read out in the fourth 1/80 second, and so on.
Therefore, the left video data and the right video data are displayed on the display device 2 in a nested manner at a total of 80 screens (high-speed images) per second, each having 40 screens per second. Since a general CRT display device has a high display and erasing speed, the left and right images can be clearly switched and displayed without leaving an afterimage.
When the image projected in this manner (the bicycle in FIG. 7) is observed simultaneously with the left and right eyes of the naked eye, it appears as a double image shifted left and right.
The reading of the left and right video data is performed by a timing signal from the timing generator 22, and this timing signal becomes the display speed on the display device 2.
[0032]
Thus, the left-eye image and the right-eye image are alternately displayed on the screen of the stereoscopic image display device 2 in time.
The glasses 4 with shutters are provided with liquid crystal shutters corresponding to the left and right liquid crystal lens units, and are opened and closed alternately in synchronism with the display of left and right video data by an infrared open / close signal R from the infrared light emitting device 3.
The brain combines these images from the left and right eyes and recognizes the 3D image on the display device 2. Since the left and right eyes respectively see the left and right screens of 40 Hz independently, it is easy to form a smooth stereoscopic image. Be recognized.
[0033]
When the left and right glasses of the glasses 4 with shutters are opened and closed in accordance with this timing signal, the left eye independently sees the left image of only the odd-numbered display screen, and the right eye independently sees the right image of only the even-numbered display screen. Stereoscopic images are generated in the human brain.
The glasses 4 with shutters and the infrared light emitting device 3 used when viewing three-dimensional (3D) images are described in detail in US Pat. No. 5,808,588, and the structure of a stereoscopic television is described in JP-A-8-20551 and JP-A-9-200804. Each of them is disclosed in detail in US Pat.
Although the case of still images has been described above, in the case of moving images, three-dimensional moving images are generated by processing still images that fluctuate little by little like a film projector and displaying them continuously as described above. And can be reproduced.
[0034]
【The invention's effect】
As described above, according to the present invention, a portion that has jumped out of the display screen and a portion that is deeper or farther from the display screen with respect to the basic position are displayed vividly. A person can recognize an extremely powerful stereoscopic image.
[Brief description of the drawings]
FIG. 1 is a diagram showing an equilibrium state of an optical axis when a human eye looks at an infinite edge.
FIG. 2 is a diagram illustrating a state of occurrence of parallax when the optical axis of a human eye is balanced.
FIG. 3 is a diagram illustrating an intersection state of optical axes when a human eye looks at a nearby object.
FIG. 4 is a diagram illustrating a state in which parallax is generated when optical axes of human eyes intersect;
FIG. 5 is a diagram illustrating various types of data included in each pixel of image data on a display screen.
FIG. 6 is a diagram showing an overall appearance of a stereoscopic video generation device of the present invention.
FIG. 7 is a diagram of a bicycle for explaining the principle of the present invention.
FIG. 8 is a diagram showing a circuit configuration of a stereoscopic video generator of the present invention.
FIG. 9 is a flowchart of a process of generating stereoscopic video emission data according to the present invention.
[Explanation of symbols]
1 Logic circuit section
2 Display device
3 Infrared light emitting device
4 glasses with shutters
5 Keyboard
6 mice
7 Changeover switch
8 Digital image generator
9 Program memory
10 Readout switch
14 Left memory
15 Right memory
16 Basic memory
17 divided memory
18 Copy memory
19 Left shift setting device
20 shift circuit
21 Right shift setting device
22 Timing generator
23 CPU
25 Distribution switch
26 Work memory
27 Video supply unit
29 storage memory

Claims (4)

By displaying the image for the left eye and the image for the right eye corresponding to the left and right eyes alternately on the electronic display screen, putting the glasses that open and close in synchronization with these displays on the left and right eyes and viewing the image In a method for generating stereoscopic video data provided to a stereoscopic video display device in which an observer recognizes a stereoscopic video,
Each pixel constituting the two-dimensional video data is given, in addition to the coordinate data, depth data having a numerical value changed from the near view to the distant view direction, and a predetermined value of the depth data is set so that the optical axes of the left and right eyes intersect. The position is set by the input device, and the central processing unit sets an area where the depth data is smaller than the basic position as a positive area, and sets the basic position where the depth data is constant as a basic area, and the depth data is larger than the basic position. The area is divided into three as a negative area, stored in a divided memory, the positive area is read out from the divided memory, and positive direction proportional data for the left or right eye is created by a shift device.
Next, according to a command from the central processing unit,
The basic region is read to create left or right eye data, and the negative region is further read to create left or right eye reverse inverse proportional data with the shift device,
These left or right eye data are stored in left or right eye memory,
The positive area is read again from the divided memory to create right or left eye positive direction proportional data with the shift device, and then the basic area is read to create right or left eye data, and further the negative area is created. A method for generating stereoscopic video data, comprising: reading out the inverse inverse proportional data for the right or left eye by the shift device; and storing the data for the right or left eye in a memory for the right or left eye, respectively. The positive direction proportional data for the left or right eye of the positive area is such that the smaller the depth data, the larger the amount of shift of the coordinate data to the right or left, and the larger the depth data, the larger the amount of shift of the coordinate data to the right or left. The data for the left or right eye of the basic area is created without a shift amount, and the inverse inverse proportional data for the left or right eye of the negative area is the left or right of the coordinate data as the depth data is smaller. The shift device creates a smaller amount of displacement, and creates a larger amount of displacement of the coordinate data to the left or right as the depth data increases,
The right or left eye positive direction proportional data of the positive area is such that the smaller the depth data, the larger the amount of shift of the coordinate data to the left or right, and the larger the depth data, the larger the amount of shift of the coordinate data to the left or right. The data for the right or left eye of the basic area is created without a shift amount, and the inverse inverse proportional data for the right or left eye of the negative area is the right or left of the coordinate data as the depth data is smaller. 2. The stereoscopic video data according to claim 1, wherein the shift device creates the amount of shift to the right or left of the coordinate data as the depth data increases, and the amount of shift to the right or left increases as the depth data increases. How it occurs. By displaying the image for the left eye and the image for the right eye corresponding to the left and right eyes alternately on the electronic display screen, putting the glasses that open and close in synchronization with these displays on the left and right eyes and viewing the image In a device that generates stereoscopic video data provided to a stereoscopic video display device in which an observer recognizes a stereoscopic video,
Each pixel constituting the two-dimensional video data is given, in addition to the coordinate data, depth data having a numerical value changed from the near view to the distant view direction, and a predetermined value of the depth data is set so that the optical axes of the left and right eyes intersect. An input device to be set as a position, an area in which the depth data is smaller than the basic position is defined as a positive area, the basic position in which the depth data is constant is defined as a basic area, and an area in which the depth data is greater than the basic position is defined as a negative area. A central processing unit that divides the data into three, a divided memory that stores each of these areas separately,
In accordance with a command from the central processing unit, the main area is read from the divided memory to create left- or right-eye proportional data, and then the basic area is read to create left- or right-eye data. While reading the negative area and creating reverse inverse proportional data for the left or right eye,
The positive area is read out again from the divided memory to create right or left eye positive proportional data, then the basic area is read out to create right or left eye data, and the negative area is read out to create a right or left eye data. Or a shift device that creates reverse inverse proportional data for the left eye,
A left or right eye memory for storing the left or right eye data,
An apparatus for generating stereoscopic video data, comprising: a right or left eye memory for storing the right or left eye data. The shift device may be configured such that the positive direction proportional data for the left or right eye of the positive region increases the shift amount of the coordinate data to the right or left as the depth data decreases, and the right or left of the coordinate data increases as the depth data increases. To the left or right eye data of the basic region is created without a displacement amount, and the reverse inverse proportional data for the left or right eye of the negative region is coordinate data as the depth data is smaller. The amount of displacement to the left or right is made smaller, and the larger the depth data, the larger the amount of displacement of the coordinate data to the left or right. The smaller the depth data, the larger the amount of shift of the coordinate data to the left or right, and the larger the depth data, the smaller the amount of shift of the coordinate data to the left or right. The data for the right or left eye of the basic area is created without a shift amount, and the inverse inverse proportional data for the right or left eye of the negative area is shifted to the right or left of the coordinate data as the depth data is smaller. 4. The three-dimensional video data generating apparatus according to claim 3, wherein the amount of displacement is reduced and the amount of displacement of the coordinate data to the right or left is increased as the depth data increases.

Images