JP2009064355A - Program, information storage medium, and image producing system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a program, etc. for enabling a flicker-free stereoscopic view image to be effectively produced. <P>SOLUTION: An image producing system includes a buffer TB having first and second areas AR1, AR2, and a drawing part for performing a drawing process for producing a multi-eye stereoscopic view image IS. The drawing part writes a first viewpoint image IR of a multi-eye stereoscopic view to the first area AR1 of the buffer TB after compression into one half in a height direction, and writes a second viewpoint image IL of the multi-eye stereoscopic view to the second area AR2 of the buffer TB after compression into one half in a height direction. Then, a multi-eye stereoscopic view image IS in which a pixel image of the first viewpoint image IR written to the first area AR1 and a pixel image of the second viewpoint image IL written to the second area AR2 are displayed alternately every line is to be produced by reading images from the buffer TB. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a program, an information storage medium, and an image generation system.

  2. Description of the Related Art Conventionally, image generation systems capable of generating stereoscopic images have attracted attention as systems for generating images with a sense of presence in the fields of games and movies. As such an image generation system, for example, there is a conventional technique disclosed in Patent Document 1. In this image generation system, a right-eye image and a left-eye image are prepared, and these images are alternately displayed on the display screen for each frame (for example, every 1/60 seconds). The viewer wears dedicated glasses, the right eye sees only the right eye image, and the left eye sees only the left eye image, thereby realizing stereoscopic viewing.

However, in this type of stereoscopic viewing, the right-eye image and the left-eye image are alternately displayed for each frame, and thus flickering occurs on the screen. Therefore, for example, when an image of a three-dimensional game is displayed in this manner, there is a problem that it is difficult to play a game for a long time.
Japanese Patent Laid-Open No. 2004-178581

  According to some aspects of the present invention, it is possible to provide a program, an information storage medium, and an image generation system that enable efficient generation of a stereoscopic image with less flicker.

  The present invention is an image generation system for multi-view stereoscopic viewing, comprising: a buffer having a first region and a second region; and a rendering unit that performs rendering processing for generating a multi-view stereoscopic image. The drawing unit compresses the first viewpoint image for multi-view stereoscopic viewing by ½ in the height direction, writes the first viewpoint image in the first area of the buffer, and outputs the second viewpoint for multi-view stereoscopic viewing. The first viewpoint image written in the first area by compressing the viewpoint image by half in the height direction, writing to the second area of the buffer, and reading the image from the buffer This pixel image and the pixel image of the second viewpoint image written in the second area relate to an image generation system that generates the multi-view stereoscopic image that is alternately displayed for each line. The present invention also relates to a program that causes a computer to function as each of the above-described units, or a computer-readable information storage medium that stores the program.

  According to the present invention, the first and second viewpoint images are compressed by half in the height direction and written to the first and second areas of the buffer. By reading out the image from this buffer, a multi-view stereoscopic image in which the first viewpoint image in the first area and the second viewpoint image in the second area are alternately displayed for each line is generated. In this way, the first and second viewpoint images are alternately displayed for each line only by the writing process compressed in the height direction of the first and second viewpoint images and the reading process of the image from the buffer. Since the generated multi-view stereoscopic image can be generated, it is possible to efficiently generate a stereoscopic image with less flicker.

  In the image generation system, the program, and the information storage medium according to the present invention, the buffer includes the first region having a width of W pixels and a height of H / 2 pixels, and a width of W pixels and a height. The second area is H / 2 pixels, the rendering unit sets the memory pitch of the buffer to 2 N bytes, the width is W pixels, and the height is H pixels. The viewpoint image is compressed in the height direction and written to the first area, and the second viewpoint image having a width of W pixels and a height of H pixels is compressed in the height direction, By writing to the area 2 and setting the memory pitch of the buffer to N bytes and reading the image from the buffer, the pixel image of the first viewpoint image and the pixel image of the second viewpoint image are written. It may generate the multiview stereoscopic image displayed alternately every.

  In this way, the pixel pitch of the first and second viewpoint images is alternated for each line by setting the memory pitch to 2N bytes when writing the image and setting the memory pitch to N bytes when reading the image. A multi-view stereoscopic image displayed on the screen can be generated.

  The image generation system, the program, and the information storage medium according to the present invention include a drawing buffer for drawing the first and second viewpoint images, and the drawing unit displays the first viewpoint image in the drawing buffer. The first viewpoint image drawn in the drawing buffer is compressed in the height direction, copied to the first area of the buffer, and the second viewpoint image is drawn in the drawing buffer. The second viewpoint image drawn in the drawing buffer may be compressed in the height direction and copied to the second area of the buffer.

  In this way, the image can be compressed in the height direction at the time of copying from the drawing buffer to the first and second areas, so that the processing can be simplified and made more efficient.

  In the image generation system, the program, and the information storage medium according to the present invention, the drawing unit texture-maps the first viewpoint image drawn in the drawing buffer to a polygon having a size of the first region. By drawing in the first area, the first viewpoint image is copied from the drawing buffer to the first area, and the second viewpoint image drawn in the drawing buffer is copied to the first area. The second viewpoint image may be copied from the drawing buffer to the second area by performing texture mapping on the polygon having the size of area 2 and drawing in the second area.

  In this way, since the image compression in the height direction can be realized only by setting the polygon size, the processing can be further simplified and made more efficient.

  In the image generation system, the program, and the information storage medium according to the present invention, the rendering unit may perform texture mapping in the height direction when texture mapping the second viewpoint image onto the polygon having the size of the second region. May be shifted by one texel.

  In this way, it is possible to generate a multi-view stereoscopic image in which the pixel images of the first and second viewpoint images are appropriately arranged.

  In the image generation system, the program, and the information storage medium according to the present invention, the drawing unit may store the image of the buffer after the first and second viewpoint images are copied to the first and second areas. You may write back to the drawing buffer.

  In this way, a multi-view stereoscopic image can be generated simply by adding a buffer or the like to an existing drawing algorithm using a drawing buffer.

  In the image generation system, the program, and the information storage medium according to the present invention, the drawing unit texture-maps the buffer image to a polygon having the size of the drawing buffer and draws the buffer in the drawing buffer. May be written back to the drawing buffer.

  In this way, it is possible to realize the writing back of the image to the drawing buffer by an efficient process that effectively uses the texture mapping.

  In the image generation system, the program, and the information storage medium according to the present invention, first and second drawing buffers are prepared as the drawing buffer, and the drawing unit is one of the first and second drawing buffers. After the image copy process from the drawing buffer to the buffer and the image write back process from the buffer to the one drawing buffer, the one drawing buffer is switched from the back buffer to the front buffer. The other different drawing buffer may be switched from the front buffer to the back buffer.

  In this way, a multi-view stereoscopic image can be generated simply by adding a buffer or the like to an existing drawing algorithm that employs a multiple buffer method.

  In the image generation system, the program, and the information storage medium according to the present invention, the drawing unit compresses the first and second viewpoint images in the height direction and writes them in the first and second areas. The pixel image of the i-th line of the first and second viewpoint images and the pixel image of the next i + 1-th line may be averaged.

  In this way, jaggy and the like can be reduced and a high-quality image can be generated.

  In the image generation system, the program, and the information storage medium according to the present invention, the drawing unit textures the first and second viewpoint images into polygons having the sizes of the first and second regions by a texel interpolation method. The averaging process of the pixel image of the i-th line and the pixel image of the i + 1-th line may be performed by mapping and drawing in the first and second regions.

  In this way, by performing texture mapping, the averaging process is automatically executed, so the process can be made more efficient.

  In the image generation system, the program, and the information storage medium according to the present invention, the drawing unit may include a pixel image of the j-th line of the first viewpoint image for the m-th line of the multi-view stereoscopic image. The pixel image of the next j + 1-th line is averaged, and the m + 1-th line of the multi-view stereoscopic image is the pixel image of the j + 1-th line of the second viewpoint image and the next j + 2-th line. The averaging process may be performed so that the line pixel images are averaged.

  In this way, it is possible to generate a multi-view stereoscopic image in which pixel images of appropriate first and second viewpoint images are displayed at appropriate positions.

  In the image generation system, the program, and the information storage medium according to the present invention, the drawing unit includes a point of sight of the virtual camera for the first viewpoint for multi-view stereoscopic viewing and the second viewpoint for the multi-view stereoscopic viewing. The first viewpoint image and the second viewpoint image may be generated by shifting the gazing point of the virtual camera in the height direction.

  In this way, a more appropriate multi-view stereoscopic image can be generated.

  Hereinafter, this embodiment will be described. In addition, this embodiment demonstrated below does not unduly limit the content of this invention described in the claim. In addition, all the configurations described in the present embodiment are not necessarily essential configuration requirements of the present invention.

1. Configuration FIG. 1 shows an example of a block diagram of an image generation system (game system) of the present embodiment. Note that the image generation system of the present embodiment may have a configuration in which some of the components (each unit) in FIG. 1 are omitted.

  The operation unit 160 is for a player to input operation data, and the function can be realized by a direction key, operation buttons, analog stick, lever, steering, accelerator, brake, microphone, touch panel display, or the like.

  The storage unit 170 serves as a work area for the processing unit 100, the communication unit 196, and the like, and its function can be realized by a RAM (DRAM, VRAM) or the like. The storage unit 170 can be configured by a volatile memory that loses data when the power is turned off, but is a storage device that is faster than the auxiliary storage device 194. Then, the game program and game data necessary for executing the game program are held in the storage unit 170.

  An information storage medium 180 (a computer-readable medium) stores programs, data, and the like, and functions thereof by an optical disk (CD, DVD), HDD (hard disk drive), memory (ROM, etc.), and the like. realizable. The processing unit 100 performs various processes of the present embodiment based on a program (data) stored in the information storage medium 180. That is, in the information storage medium 180, a program for causing a computer (an apparatus including an operation unit, a processing unit, a storage unit, and an output unit) to function as each unit of the present embodiment (a program for causing the computer to execute processing of each unit). Is memorized.

  The display unit 190 outputs an image generated according to the present embodiment, and its function can be realized by a CRT, LCD, touch panel display, HMD (head mounted display), or the like. The sound output unit 192 outputs the sound generated by the present embodiment, and its function can be realized by a speaker, headphones, or the like.

  The auxiliary storage device 194 (auxiliary memory, secondary memory) is a large-capacity storage device that is used to supplement the capacity of the storage unit 170, and may be a memory card such as an SD memory card or a multimedia card, an HDD, or the like. realizable. The auxiliary storage device 194 is detachable, but may be built-in. The auxiliary storage device 194 is used to save save data such as the game midway results, personal image data and music data of the player (user), and the like.

  The communication unit 196 communicates with the outside (for example, another image generation system, a server, or a host device) via a wired or wireless network, and functions as a communication ASIC, a communication processor, or the like. It can be realized by hardware and communication firmware.

  Note that a program (data) for causing a computer to function as each unit of the present embodiment is obtained from an information storage medium included in a server (host device) via an information storage medium 180 (or storage unit 170, auxiliary storage) via a network and communication unit 196. May be distributed to the device 194). Use of an information storage medium by such a server (host device) can also be included in the scope of the present invention.

  The processing unit 100 (processor) performs game processing, image generation processing, sound generation processing, and the like based on operation data from the operation unit 160, a program, and the like. The processing unit 100 performs various processes using the storage unit 170 (main storage unit 172) as a work area. The functions of the processing unit 100 can be realized by hardware such as various processors (CPU, GPU, etc.), ASIC (gate array, etc.), and programs.

  The processing unit 100 includes a game calculation unit 102, an object space setting unit 104, a moving body calculation unit 106, a virtual camera control unit 108, a drawing unit 120, and a sound generation unit 130. Note that some of these may be omitted.

  The game calculation unit 102 performs game calculation processing. Here, as a game calculation, a process for starting a game when a game start condition is satisfied, a process for advancing the game, a process for calculating a game result, or a process for ending a game when a game end condition is satisfied and so on.

  The object space setting unit 104 displays various display objects such as model objects (moving bodies such as cars, fighters, people, robots, missiles, and bullets), maps (terrain), buildings, courses (roads), trees, and walls. Processing for setting an object (an object composed of a primitive surface such as a polygon, a free-form surface, or a subdivision surface) in the object space is performed. In other words, the position and rotation angle of the object in the world coordinate system (synonymous with direction and direction) are determined, and the rotation angle (rotation angle around the X, Y, and Z axes) is determined at that position (X, Y, Z). Arrange objects. Specifically, the model data storage unit MS of the storage unit 170 stores object data that is data such as the position, rotation angle, movement speed, and movement direction of the object in association with the object number. Then, the object data is sequentially updated by a moving object calculation process or the like of the moving object calculation unit 106.

  The object space setting unit 104 includes a culling processing unit 105. The culling processing unit 105 performs culling processing by software, for example. Specifically, a process of deleting an object that is not displayed in the frame from the object list of the object data is performed. For example, an object that does not exist within the field of view of the virtual camera in the frame and does not need to be displayed is deleted in advance. In this way, it is not necessary to create vertex data or the like for the polygons constituting the deleted object, so that the processing load can be greatly reduced.

  The moving object calculation unit (moving object control unit) 106 performs an operation for moving the moving object (moving object). Also, calculations for operating the moving body are performed. That is, based on operation data input by the player through the operation unit 160, a program (movement / motion algorithm), various data (motion data), etc., a moving body (object, model object) is moved in the object space, Performs processing to move the moving body (motion, animation). Specifically, a simulation process for sequentially obtaining movement information (position, rotation angle, speed, or acceleration) and motion information (part object position or rotation angle) of a moving body for each frame (1/60 second). I do. A frame is a unit of time for performing a moving / movement process (simulation process) and an image generation process of a moving object.

  More specifically, the moving body calculation unit 106 performs processing for reproducing the motion of the object based on the motion data stored in the motion data storage unit MTS. That is, motion data including the position or rotation angle (direction) of each part object (bone constituting the skeleton) constituting the object (skeleton) is read from the motion data storage unit MTS. Then, by moving each part object (bone) of the object (by deforming the skeleton shape), the motion of the object is reproduced.

  The virtual camera control unit 108 performs a virtual camera (viewpoint) control process for generating an image that can be seen from a given (arbitrary) viewpoint in the object space. Specifically, a process for controlling the position (X, Y, Z) or the rotation angle (rotation angle about the X, Y, Z axis) of the virtual camera (a process for controlling the viewpoint position, the line-of-sight direction, or the angle of view). I do.

  For example, when a moving body such as a car, character, or fighter is photographed from behind using a virtual camera, the position or rotation angle of the virtual camera (virtual camera is set so that the virtual camera follows changes in the position or rotation of the moving body. The direction). In this case, the virtual camera can be controlled based on information such as the position, rotation angle, or speed of the moving object obtained by the moving object computing unit 106. Alternatively, the virtual camera may be controlled to rotate at a predetermined rotation angle or to move along a predetermined movement path. In this case, the virtual camera is controlled based on the virtual camera data for specifying the position (movement path) or rotation angle of the virtual camera.

  In the present embodiment, the virtual camera control unit 108 controls the virtual cameras of a plurality of viewpoints in order to generate a multi-view stereoscopic image. Specifically, the binocular parallax or the like for controlling the first viewpoint (for example, right eye) virtual camera and the second viewpoint (for example, left eye) virtual camera to make the player feel a stereoscopic effect, etc. Realize.

  The drawing unit 120 (image generation unit) performs drawing processing based on the results of various processing (game processing and simulation processing) performed by the processing unit 100, thereby generating an image and outputting it to the display unit 190. In order to generate a so-called three-dimensional game image, first, vertex data (vertex position coordinates, texture coordinates, color data, normal vector, α value, etc.) of each vertex of the model (object) is generated. Based on the above, vertex processing (shading by a vertex shader) is performed. When performing the vertex processing, vertex generation processing (tessellation, curved surface division, polygon division) for re-dividing the polygon may be performed as necessary.

  In vertex processing (vertex shader processing), vertex movement processing, coordinate transformation (world coordinate transformation, camera coordinate transformation), clipping processing, perspective transformation, etc., according to a vertex processing program (vertex shader program, first shader program) Geometry processing is performed, and based on the processing result, the vertex data given to the vertex group constituting the object is changed (updated, adjusted). Then, rasterization (scan conversion) is performed based on the vertex data after the vertex processing, and the surface of the polygon (primitive) is associated with the pixel. Subsequent to rasterization, pixel processing (shading or fragment processing by a pixel shader) for drawing pixels (fragments forming a display screen) constituting an image is performed.

  In pixel processing (pixel shader processing), according to the pixel processing program (pixel shader program, second shader program), various processes such as texture reading (texture mapping), color data setting / changing, translucent composition, anti-aliasing, etc. The final drawing color of the pixels constituting the image is determined, and the drawing color of the perspective-converted model is stored in the drawing buffer DB (buffer that can store image information in units of pixels; VRAM, rendering target, frame buffer). Output (draw). That is, in pixel processing, per-pixel processing for setting or changing image information (color, normal, luminance, α value, etc.) in units of pixels is performed. Thereby, an image that can be seen from the virtual camera (given viewpoint) in the object space is generated.

  Note that the vertex processing and pixel processing are realized by hardware that enables polygon (primitive) drawing processing to be programmed by a shader program written in a shading language, so-called programmable shaders (vertex shaders and pixel shaders). Programmable shaders can be programmed with vertex-level processing and pixel-level processing, so that the degree of freedom of drawing processing is high, and expressive power is greatly improved compared to conventional hardware-based fixed drawing processing. Can do.

  The drawing unit 120 includes a lighting processing unit 122 and a texture mapping unit 124. Note that some of these may be omitted.

  The lighting processing unit 122 performs lighting processing (shading processing) based on an illumination model or the like. Specifically, this lighting processing includes light source information (light source vector, light source color, brightness, light source type, etc.), line-of-sight vector of virtual camera (viewpoint), normal vector of object (translucent object), object material ( Color, material) and the like. Illumination models include the Lumbard diffuse lighting model that considers only ambient light and diffuse light, the phone lighting model that considers specular light in addition to ambient light and diffuse light, and the Bling phone lighting model. is there.

  The texture mapping unit 124 performs processing for mapping a texture to an object (polygon). Here, the texture mapping process is a process of mapping a texture (texel value) stored in the texture storage unit TS to an object. Specifically, the texture (surface properties such as color and α value) is read from the texture storage unit TS using the texture coordinates set (given) to the vertices and pixels of the object (primitive surface). Then, a texture that is a two-dimensional image or pattern is mapped to the object. In this case, processing for associating pixels with texels, bilinear interpolation (texel interpolation in a broad sense), and the like are performed.

  In the present embodiment, the drawing unit 120 performs a drawing process for generating a multi-view stereoscopic image (n-eye stereoscopic image, n ≧ 2). Specifically, the drawing unit 120 compresses (reduces), for example, a first viewpoint image (for example, a right-eye image) for multi-view stereoscopic viewing (two-view stereoscopic viewing) to, for example, 1/2 in the height direction (vertical direction). Thus, the data is written into the first area AR1 (for example, the left half area) of the buffer TB. Further, the second viewpoint image (for example, the left-eye image) for multi-view stereoscopic viewing is compressed (reduced) to, for example, 1/2 in the height direction (vertical direction), and the second area AR2 (for example, the right half) of the buffer TB Side area). Then, by reading out the image (the image of AR1, AR2) from the buffer TB, the pixel image (R, G, B, α) of the first viewpoint image written in the first area AR1 and the second area AR2 are read. A multi-view stereoscopic image is generated in which the pixel image of the written second viewpoint image is alternately displayed for each line (for example, for each scanning line or for each bitmap line).

  More specifically, the first and second areas AR1 and AR2 are areas having a width of W pixels (for example, 1920 pixels) and a height of H / 2 pixels (for example, 1080/2 = 540 pixels). Yes.

  At the time of image writing, the memory pitch of the buffer TB is set to 2N bytes (for example, 2 × 4 bytes × 2048 pixels = 16384 bytes). Here, the memory pitch (pitch size) is a distance in bytes between the address representing the head of the bitmap line and the address representing the head of the next line. In the buffer TB set in this way, the drawing unit 120 writes the first viewpoint image having a width of W pixels and a height of H pixels in the first area AR1. A second viewpoint image having a width of W pixels and a height of H pixels is written in the second area AR2. For example, the first viewpoint image is drawn in the drawing buffer DB (DB1 or DB2), and the drawn first viewpoint image is compressed in the height direction and copied to the first area AR1 of the buffer TB. In addition, the second viewpoint image is drawn in the drawing buffer DB (DB1 or DB2), the drawn second viewpoint image is compressed in the height direction, and is copied to the second area AR2 of the buffer TB. In this case, the copy processing is performed by mapping the first or second viewpoint image as a texture to the polygons of the sizes of the first and second areas AR1 and AR2 (virtual polygons not displayed on the screen). This can be realized by drawing a polygon in the first or second area AR1, AR2.

  On the other hand, when reading an image, the memory pitch of the buffer TB is set to N bytes. The drawing unit 120 generates a multi-view stereoscopic image by reading an image from the buffer TB set in this way. For example, after the first and second viewpoint images are copied to the first and second areas AR1 and AR2, the image in the buffer TB is written back (written back) to the drawing buffer DB (DB1 or DB2). The write-back process in this case can be realized by texture mapping the image in the buffer TB to a polygon having the size of the drawing buffer DB (DB1 or DB2) and drawing the polygon in the drawing buffer DB.

  The sound generation unit 130 performs sound processing based on the results of various processes performed by the processing unit 100, generates game sounds such as BGM, sound effects, or sounds, and outputs the game sounds to the sound output unit 192.

  Note that the image generation system of the present embodiment may be a system dedicated to the single player mode in which only one player can play, or may be a system having a multiplayer mode in which a plurality of players can play. Further, when a plurality of players play, game images and game sounds to be provided to the plurality of players may be generated using one terminal, or connected via a network (transmission line, communication line) or the like. Alternatively, it may be generated by distributed processing using a plurality of terminals (game machine, mobile phone).

2. 2. Method of the Present Embodiment 2.1 Stereoscopic Method First, an example of the stereoscopic method employed in the present embodiment will be described with reference to FIGS. 2 and 3.

  The multi-view stereoscopic image generated by the image generation system 300 (for example, a game device) in FIG. 2 is displayed on the display screen of the display device 310 (for example, a high-definition television). A film 312 in which right-eye polarizing filters and left-eye polarizing filters are alternately arranged for each scanning line is attached to the display screen. The player views a stereoscopic image by wearing the polarizing glasses 320 corresponding to these polarizing filters and viewing the display screen.

  Specifically, the right eye image and the left eye image are alternately displayed on the display screen of the display device 310 for each scanning line. The player views the right-eye image of each scanning line polarized by the right-eye polarizing filter of the film 312 via the polarizing filter provided on the right-eye portion 322 of the polarizing glasses 320. Further, the left eye image of each scanning line polarized by the left eye polarizing filter of the film 312 is viewed through the left eye portion 324 of the polarizing glasses 320.

  2, the player can view the right-eye image and the left-eye image simultaneously through the right-eye portion 322 and the left-eye portion 324 of the polarizing glasses 320. Therefore, it is possible to realize an ideal stereoscopic display with less flickering and less fatigue compared to a conventional stereoscopic method in which a right-eye image and a left-eye image are alternately displayed for each frame. In particular, a fast-moving video such as a three-dimensional game video has disadvantages such as loss of stereoscopic effect and increased eye fatigue in the conventional stereoscopic method, but according to the stereoscopic method shown in FIG. Such a drawback can be solved.

  FIG. 3 conceptually shows a right-eye image IR, a left-eye image IL, and a multi-view stereoscopic image IS. In the multi-view stereoscopic image IS of FIG. 3, the pixel image of the right-eye image IR is displayed on the odd-numbered scan lines, and the pixel image of the left-eye image IL is displayed on the even-numbered scan lines. The player views the pixel image of the odd-numbered scan line through the right eye portion 322 of the polarizing glasses 320 and also views the pixel image of the even-numbered scan line through the left eye portion 324, thereby obtaining a stereoscopic image. I can enjoy it.

  4A, 4B, and 5 show examples of the right-eye image IR, the left-eye image IL, and the multi-view stereoscopic image IS that are generated by the image generation system of the present embodiment. These are examples of images when the method of this embodiment is applied to a three-dimensional race game.

  4A and 4B, pixel images of the right-eye image IR and the left-eye image IL are generated for all scanning lines. For example, in the case of full high-definition support, pixel images are generated for all 1080 scan lines. Then, the multi-view stereoscopic image IS of FIG. 5 is generated by combining the right-eye image IR and the left-eye image IL that are compatible with full high-definition. Accordingly, since the stereoscopic game video can be displayed at full high-definition resolution, the virtual reality of the player can be dramatically improved. That is, in the conventional stereoscopic method, there is a problem that the image quality is deteriorated because the resolution is lowered or the frame rate is lowered by making the stereoscopic image, but FIGS. This method can solve such a problem.

2.2 Stereoscopic Image Generation Method Next, a stereoscopic image generation method according to the present embodiment will be described. In the present embodiment, the method described below is adopted in order to generate a high-quality stereoscopic image as shown in FIGS. In the following, a case where the first viewpoint image is a right-eye image and the second viewpoint image is a left-eye image will be described as an example, but the present embodiment is not limited to this. For example, the first viewpoint image may be a left eye image or the like, and the second viewpoint image may be a right eye image or the like.

  As shown in FIG. 6, in this embodiment, first and second areas AR1 and AR2 are secured in the buffer TB. Then, a right-eye image IR (first viewpoint image in a broad sense) as illustrated in FIG. 4A is generated, and the right-eye image IR is compressed (reduced) in the height direction (vertical direction). And write to the first area AR1 of the buffer TB. Further, a left-eye image IL (second viewpoint image in a broad sense) as illustrated in FIG. 4B is generated, the left-eye image IL is compressed in the height direction, and the second eye of the buffer TB is generated. Write to area AR2.

  Then, by reading out the image from the buffer TB, a multi-view stereoscopic image IS as exemplified in FIG. 5 is generated (synthesized). That is, a multi-view stereoscopic image IS in which the right-eye pixel image in the area AR1 and the left-eye pixel image in the area AR2 are alternately displayed for each line (for each scanning line and for each bitmap line) is generated.

  Specifically, the right-eye image IR and the left-eye image IL are images having a width of W pixels and a height of H pixels. In the case of full high-definition support, for example, W = 1920 and H = 1080. On the other hand, the areas AR1 and AR2 are areas having a width of W pixels and a height of H / 2 pixels.

  At the time of writing an image to the buffer TB, the memory pitch of the buffer TB (distance in bytes from the head address of the next line to the head address of the next line) is set to 2N bytes. For example, when one pixel (R, G, B, α) is 4 bytes, for example, 2N bytes = 2 × 4 × 2048 bytes for full high-definition support. In FIG. 6, the right-eye image IR having a width of W pixels and a height of H pixels is compressed by half in the height direction and written in the area AR1, and the width is W pixels and the height is The left-eye image IL that is H pixels is compressed in half in the height direction and written in the area AR2.

  On the other hand, when reading an image from the buffer TB, the memory pitch of the buffer TB is set to N bytes. Then, by reading the right-eye image IR and the left-eye image IL from the buffer TB set in this way, a multi-view stereoscopic image IS as illustrated in FIG. 5 is synthesized.

  Specifically, by setting the memory pitch (pitch size) set to 2N bytes at the time of writing to N bytes at the time of reading, the right-eye pixel image of the first line in the area AR1 is a multi-view stereoscopic image. The left-eye pixel image of the first line in the area AR2 displayed on the first line of the IS is displayed on the second line of the multi-view stereoscopic image IS. The right-eye pixel image of the second line in the area AR1 is displayed on the third line of the multi-view stereoscopic image IS, and the left-eye pixel image of the second line in the area AR2 is displayed in the multi-view stereoscopic image IS. Is displayed in the fourth line. That is, the right-eye pixel image and the left-eye pixel image are alternately displayed for each line.

2.3 Use of Rendering Buffer The right-eye image IR and the left-eye image IL generated by rendering may be directly rendered in the areas AR1 and AR2 of the buffer TB, but FIG. 7 (A) and FIG. ), After drawing in the drawing buffer DB, it may be copied to the areas AR1 and AR2.

  That is, in FIG. 7A, the right-eye image IR is drawn in the drawing buffer DB, the drawn right-eye image IR is compressed in the height direction, and an area AR1 of the buffer TB that functions as a temporary buffer (work buffer). Copy to. In FIG. 7B, the left-eye image IL is drawn in the drawing buffer DB, the drawn left-eye image IL is compressed in the height direction, and copied to the area AR2 of the buffer TB.

  In this way, compression (averaging, thinning, reduction) of the image in the height direction may be performed at the time of copy processing from the drawing buffer DB to the areas AR1 and AR2, and image compression and right eye / left eye are performed. Since it is not necessary to simultaneously generate the image for processing, the process can be simplified and improved in efficiency.

  In this case, the copy processing in FIGS. 7A and 7B can be realized by using, for example, texture mapping.

  Specifically, as shown in FIG. 8A, the right-eye image IR drawn in the drawing buffer DB is texture-mapped to a polygon PL (virtual primitive surface not displayed on the screen) having the size of the area AR1. In this case, the texture mapping is set not to the point sampling method but to, for example, a texel interpolation method (bilinear filter mode in a narrow sense). Then, the polygon PL in which the right-eye image IR is mapped as a texture is rendered in the area AR1 in this way, thereby realizing the copy process of FIG.

  Further, as shown in FIG. 8B, the left eye image IL drawn in the drawing buffer DB is texture-mapped to the polygon PL having the size of the area AR2. In this case, texture mapping is set to the texel interpolation method. Also, texture mapping is performed by shifting texture coordinates (for example, V coordinates in the UV space) in the height direction by, for example, one texel. Then, the polygon PL in which the left eye image IL is mapped as a texture is rendered in the area AR2, thereby realizing the copy process of FIG.

  If copy processing is realized by texture mapping as shown in FIGS. 8A and 8B, image compression in the height direction can be realized only by setting the size of the polygon PL, thereby simplifying and increasing the efficiency of the processing. Can be planned. Moreover, since the averaging process of pixel images of adjacent lines as will be described later can be realized using a texture mapping texel interpolation method (bilinear filter), the processing can be made more efficient.

  As shown in FIGS. 7A and 7B, after the right-eye image IR and the left-eye image IL are copied to the areas AR1 and AR2, the image in the buffer TB is written back to the drawing buffer DB. Specifically, the memory pitch of the buffer TB is changed from 2N bytes to N bytes, and the image of the buffer TB is written back to the drawing buffer DB to generate a multi-view stereoscopic image IS as illustrated in FIG.

  In this way, the temporary buffer of the right-eye image IR and the left-eye image IL can be simply added to the existing drawing algorithm using the drawing buffer DB by adding a buffer TB as a temporary buffer, a copy / write-back algorithm, or the like. Copy processing to the buffer TB is realized, and a multi-view stereoscopic image IS can be generated.

  Note that image write-back (write back) to the drawing buffer DB can be realized by a technique using texture mapping in the same manner as in FIGS. 8A and 8B. Specifically, the pitch size of the buffer TB is changed from 2N bytes to N bytes. Then, the texture of the image of the buffer TB (the entire image of TB) is texture-mapped to a polygon having the size of the drawing buffer DB (DB1 or DB2). Then, by drawing the texture-mapped polygon of the image in the buffer TB in the drawing buffer DB, the image in the buffer TB is written back into the drawing buffer DB.

2.4 Double Buffer Method When the double buffer method (or a multiple buffer method such as a triple buffer) is used as the drawing buffer DB, it is desirable to adopt the method shown in FIGS. 9A and 9B. .

  For example, in FIG. 9A, first and second drawing buffers DB1 and DB2 are prepared (reserved) as the drawing buffer DB. 9A, the first drawing buffer DB1 is set as a back buffer (drawing buffer), and the second drawing buffer DB2 is set as a front buffer (display buffer). In the multiple buffer system such as a triple buffer, three or more drawing buffers including the first and second drawing buffers DB1 and DB2 may be prepared (reserved).

  In FIG. 9A, the drawing buffer DB1 (one drawing buffer of the first and second drawing buffers in a broad sense) to the buffer TB is described with reference to FIGS. 7A and 7B. The copied image is processed. In addition, an image write-back process from the buffer TB to the drawing buffer DB1 is performed. As a result, a multi-view stereoscopic image IS to be displayed in the next frame is generated in the drawing buffer DB1.

  After the multi-view stereoscopic image IS is generated in the drawing buffer DB1, the drawing buffer DB1 is switched from the back buffer to the front buffer (flip) as shown in FIG. 9B. Thereby, the multi-view stereoscopic image IS of the drawing buffer DB1 is displayed on the screen of the display device. At this time, the drawing buffer DB2 (in other words, the other drawing buffer different from one) is switched from the front buffer to the back buffer.

  In FIG. 9B, the image copy processing described in FIGS. 7A and 7B is performed from the drawing buffer DB2 to the buffer TB. Further, an image write-back process from the buffer TB to the drawing buffer DB2 is performed. As a result, a multi-view stereoscopic image IS to be displayed in the next frame is generated in the drawing buffer DB2.

  According to the methods shown in FIGS. 9A and 9B, a buffer TB that is a temporary buffer, a copy / write-back algorithm, and the like are added to the existing drawing algorithm of the double buffer method. The double buffer method and the method for generating the multi-view stereoscopic image IS according to the present embodiment can be compatible.

2.5 Averaging Processing When compression processing in the height direction (scanning line arrangement direction) is performed as shown in FIG. 6, averaging processing (interpolation processing) of pixel images of adjacent lines may be performed. desirable.

  Specifically, as shown in FIG. 10, when the right-eye image IR (first viewpoint image) is compressed in the height direction and written in the area AR1, the i-th (i is a natural number) line of the right-eye image. And the pixel image of the next (i + 1) th line are averaged. Similarly, when the left eye image IL (second viewpoint image) is compressed in the height direction and written in the area AR2, the pixel image of the i-th line and the pixel image of the next i + 1-th line of the left-eye image are written. Averaging processing is performed.

  For example, in FIG. 10, the pixel images of the first, second, third, and fourth lines (RGB images of specific pixels in the line) of the right-eye image IR are R1, R2, R3, and R4. In this case, the pixel image of the first line in the area AR1 becomes (R1 + R2) / 2 by the averaging process of R1 and R2, and the pixel image of the second line becomes the averaging process of R3 and R4. (R3 + R4) / 2. The same applies to the other lines of the right eye image IR and the left eye image IL.

  By performing such an averaging process, even when the right-eye pixel image and the left-eye pixel image are alternately displayed for each line in a multi-view stereoscopic image, for example, a high-definition resolution compatible with, for example, high-definition is maintained. The stereoscopic image thus displayed can be displayed. In addition, the occurrence of jaggy and aliasing can be reduced, and a high-quality stereoscopic image can be generated.

  In this averaging process, as described in FIGS. 8A and 8B, the texture mapping mode is set to the texel interpolation method (bilinear filter mode), and the texture is set in the height direction. This can be done automatically by performing compressed texture mapping. Specifically, the right-eye image IR and the left-eye image IL are texture-mapped to the polygons of the sizes of the areas AR1 and AR2 by the texel interpolation method, and are drawn in the areas AR1 and AR2. And the pixel image of the (i + 1) th line are averaged.

  In other words, when the texture is compressed by half in the height direction, the texture texel intervals of the right-eye image IR and the left-eye image IL do not match the pixel intervals of the areas AR1 and AR2, and the texel interval is, for example, 1 pixel interval. / Doubled. Therefore, an image obtained by averaging a plurality of texel images is mapped to each pixel in the areas AR1 and AR2. Thereby, since the averaging process as shown in FIG. 10 is automatically executed, the process can be made efficient.

  FIG. 11 shows a more detailed specific example of the averaging process of the present embodiment. In FIG. 11, for the m-th (m is a natural number) line of the multi-view stereoscopic image IS, the pixel image of the j-th (j is a natural number) line of the right-eye image IR (first viewpoint image) and the next The pixel image of the (j + 1) th line is averaged. For the (m + 1) th line of the multi-view stereoscopic image IS, the pixel image of the j + 1th line and the pixel image of the next j + 2 line of the left eye image IL (second viewpoint image) are averaged. . This can be realized by shifting the texture coordinates by one texel described with reference to FIG.

  For example, assume that the pixel images of the first to sixth lines of the right-eye image IR are R1 to R6, and the pixel images of the first to seventh lines of the left-eye image IL are L1 to L7. In this case, the first line (m-th line) of the multi-view stereoscopic image IS includes the pixel image R1 of the first line (j-th line) of the right-eye image IR and the second line ( An image (R1 + R2) / 2 obtained by averaging pixel images R2 of the (j + 1) th line) is displayed (arranged). The second line (m + 1th line) of the multi-view stereoscopic image IS includes the pixel image L2 of the second line (j + 1th line) of the left-eye image IL and the third line (j + 2th line). An image (L2 + L3) / 2 obtained by averaging the pixel image L3 of (line) is displayed.

  Similarly, the third line of the multi-view stereoscopic image IS has an image (R3 + R4) / 2 obtained by averaging the pixel image R3 of the third line of the right-eye image IR and the pixel image R4 of the fourth line. Is displayed. Further, the fourth line of the multi-view stereoscopic image IS displays an image (L4 + L5) / 2 obtained by averaging the pixel image L4 of the fourth line of the image for left eye IL and the pixel image L5 of the fifth line. Is done. The same applies to the other lines. In FIG. 11, the pixel image L1080 is displayed in the final line of the multi-view stereoscopic image IS because there is no pixel image to be synthesized with the pixel image L1080.

  According to the method shown in FIG. 11, the right-eye pixel image or the left-eye pixel image is displayed at an appropriate position in the multi-view stereoscopic image IS, so that jaggy and anti-aliasing can be effectively reduced.

  For example, FIG. 12 shows an example of an image generated when the averaging method of this embodiment is adopted. FIG. 12 is an enlarged image of a guardrail portion in a racing game, for example. As shown in FIG. 12, since the display of blurring due to the averaging process is performed on the edge portion of the guardrail that changes in a staircase pattern, jaggies become inconspicuous and anti-aliasing can be reduced.

  On the other hand, FIG. 13 is an example of an image generated when the averaging method of the present embodiment is not adopted. Specifically, in FIG. 13, the right-eye image R1 is displayed on the first line of the multi-view stereoscopic image IS, the left-eye image L2 is displayed on the second line, the right-eye image R3 is displayed on the third line, and the fourth line. The left eye image L4 is displayed on the line. In FIG. 13, since the blur as shown in FIG. 12 is not applied to the edge portion of the guard rail that changes in a staircase pattern, jaggies are conspicuous and the image quality deteriorates. According to the averaging method of this embodiment, such a problem can be solved.

  In FIG. 11, the averaging process is not performed on the final line (L1080) of the multi-view stereoscopic image IS, and there is a problem that correct display cannot be realized in this portion. In order to solve such a problem, a method of shifting the gazing point (drawing range) of the virtual camera may be employed.

  Specifically, as shown in FIG. 14, the gazing point of the virtual camera VCR for the right eye (for the first viewpoint) for multi-view stereoscopic viewing and the virtual for the left eye (for the second viewpoint) for multi-view stereoscopic viewing. The gazing point of the camera VCL is shifted (shifted), for example, in the height direction (the direction in which the scanning lines are arranged) to generate a right-eye image and a left-eye image (first viewpoint image, second viewpoint image). ing.

  More specifically, the gazing point of the virtual camera VCR for the right eye is shifted by, for example, 0.5 pixels upward, and the gazing point of the virtual camera VCL for the left eye is, for example, only 0.5 pixels downward. Shift. As a result, the left and right gazing points are shifted up and down by, for example, one pixel. As a result, the average of the images L1079 and L1080 is also displayed on the final line in FIG. 11, and the quality of the generated image can be further improved.

2.6 Detailed Processing Example Next, a detailed processing example of this embodiment will be described with reference to the flowcharts of FIGS. 15 and 16.

  First, as described with reference to FIGS. 6, 9A, and 9B, the first and second drawing buffers DB1 having a memory pitch of N bytes, a width of W pixels, and a height of H pixels, DB2 is secured (step S1). Further, a buffer TB having first and second areas AR1 and AR2 having a width of W pixels and a height of H / 2 pixels, a memory pitch of 2N bytes, and a height of H / 2 pixels is secured. (Step S2).

  Next, the right-eye image IR that can be seen from the right-eye virtual camera VCR is drawn in the first drawing buffer DB1 (step S3). Then, as described with reference to FIGS. 7A to 8B, the first area AR1 (W × H / 2) is obtained using the image (W × H pixel) drawn in the first drawing buffer DB1 as a texture. By drawing a polygon in (pixel), the image of DB1 is compressed in the height direction and copied to AR1. At this time, the entire area of the image in the first drawing buffer DB1 is mapped to the polygon to be written in the first area AR1 (step S4).

  Next, the left-eye image IL that can be seen from the left-eye virtual camera VCL is drawn in the first drawing buffer DB1 (step S5). Then, the image drawn in the first drawing buffer DB1 (W × H pixel) is used as a texture, and a polygon is drawn in the second area AR2 (W × H / Z pixel), thereby heightening the image in DB1. Compress in direction and copy to AR2. At this time, the polygon written in the second area AR2 is mapped by shifting the image of the first drawing buffer DB1 downward by 1 texel (step S6).

  Next, the entire area of the buffer TB is rendered as a texture having a memory pitch of N bytes, a width of W pixels, and a height of H pixels, and polygons are drawn over the entire area of DB1 (W × H pixels). Is copied to the first drawing buffer DB1 (step S7). At the next VSYNC timing (frame switching timing), the first drawing buffer DB1 is switched from the back buffer to the front buffer as described with reference to FIGS. 9A and 9B (step S8).

  Next, the right-eye image IR that can be seen from the right-eye virtual camera VCR is drawn in the second drawing buffer DB2 (step S9). Then, the image drawn in the second drawing buffer DB2 (W × H pixel) is used as a texture, and a polygon is drawn in the first area AR1 (W × H / 2 pixel), thereby heightening the image in DB2. Compress in direction and copy to AR1. At this time, the entire area of the image in the second drawing buffer DB2 is mapped to the polygon to be written in the first area AR1 (step S10).

  Next, the left-eye image IL that can be seen from the left-eye virtual camera VCL is drawn in the second drawing buffer DB2 (step S11). Then, the image drawn in the second drawing buffer DB2 (W × H pixels) is used as a texture, and a polygon is drawn in the second area AR2 (W × H / 2 pixels), thereby heightening the image in DB2. Compress in direction and copy to AR2. At this time, the polygon written in the second area AR2 is mapped by shifting the image in the second drawing buffer DB2 downward by 1 texel (step S12).

  Next, by rendering the entire area of the buffer TB as a texture having a memory pitch of N bytes, a width of W pixels, and a height of H pixels, a polygon is drawn over the entire area (W × H pixels) of the second drawing buffer DB2. The image in the buffer TB is copied to the second drawing buffer DB2 (step S13).

  At the next VSYNC timing, the second drawing buffer DB2 is switched from the back buffer to the front buffer (step S14).

3. Hardware Configuration FIG. 17A shows a hardware configuration example capable of realizing the present embodiment.

  The CPU 900 (main processor) is a multi-core processor including a plurality of CPU cores 1, CPU cores 2, and CPU cores 3. The CPU 900 includes a cache memory (not shown). Each of the CPU cores 1, 2, and 3 is provided with a vector calculator and the like. Each of the CPU cores 1, 2, and 3 can execute, for example, two H / W thread processes in parallel without performing a context switch, and a multi-thread function is supported by hardware. A total of three CPU cores 1, 2, and 3 can execute six H / W thread processes in parallel.

  The GPU 910 (drawing processor) performs vertex processing and pixel processing to realize drawing (rendering) processing. Specifically, according to the shader program, the vertex data is created / changed and the drawing color of the pixel (fragment) is determined. When an image for one frame is written into the VRAM 920 (frame buffer), the image is displayed on a display such as a TV via a video output. The main memory 930 functions as a work memory for the CPU 900 and the GPU 910. Further, in the GPU 910, a plurality of vertex threads and a plurality of pixel threads are executed in parallel, and a multi-thread function of drawing processing is supported by hardware. The GPU 910 is also provided with a hardware tessellator. The GPU 910 is a unified shader type in which the vertex shader and the pixel shader are not distinguished in terms of hardware.

  The bridge circuit 940 (south bridge) is a circuit that controls the flow of information inside the system, and incorporates a controller such as a USB controller (serial interface), a network communication controller, an IDE controller, or a DMA controller. The bridge circuit 940 implements an interface function among the game controller 942, the memory card 944, the HDD 946, and the DVD drive 948.

  Note that the hardware configuration capable of realizing the present embodiment is not limited to FIG. 17A, and may be, for example, a configuration as shown in FIG.

  In FIG. 17B, the CPU 902 includes a processor element PP and eight processor elements PE1 to PE8. The processor element PP is a general-purpose processor core, and the processor elements PE1 to PE8 are processor cores having a relatively simple configuration. The architectures of the processor element PP and the processor elements PE1 to PE8 are different, and the processor elements PE1 to PE8 are SIMD type processor cores that can simultaneously perform the same processing on a plurality of data with one instruction. Thereby, multimedia processing such as streaming processing can be executed efficiently. The processor element PP can execute two H / W thread processes in parallel, and each of the processor elements PE1 to PE8 can execute one H / W thread process. Therefore, the CPU 902 can execute a total of 10 H / W thread processes in parallel.

  In FIG. 17B, the GPU 912 and the CPU 902 are closely linked, and the GPU 912 can directly perform the rendering process on the main memory 930 on the CPU 902 side. Further, for example, the CPU 902 can perform geometry processing to transfer vertex data, or easily return data from the GPU 912 to the CPU 902. It is also easy for the CPU 902 to perform rendering pre-processing and post-processing, and the CPU 902 can execute tessellation (plane division) and dot fill. For example, the CPU 902 can perform a process with a high level of abstraction, and the GPU 912 can perform a detailed process with a low level of abstraction.

  When the processing of each unit of the present embodiment is realized by hardware and a program, a program for causing the hardware (computer) to function as each unit of the present embodiment is stored in the information storage medium. More specifically, the program instructs a processor (CPU, GPU), which is hardware, to pass data if necessary. And a processor implement | achieves the process of each part of this invention based on the instruction | indication and the passed data.

  Although the present embodiment has been described in detail as described above, it will be easily understood by those skilled in the art that many modifications can be made without departing from the novel matters and effects of the present invention. Accordingly, all such modifications are intended to be included in the scope of the present invention. For example, in the specification or drawings, terms (for the right eye) described at least once together with different terms (first viewpoint image, second viewpoint image, first viewpoint, second viewpoint, etc.) having a broader meaning or the same meaning Image, left eye image, right eye, left eye, etc.) may be replaced by the different terms anywhere in the specification or drawings.

  Also, the drawing processing, copy processing, write-back processing, texture mapping processing, drawing buffer, buffer securing method, and the like of the first and second viewpoint images are not limited to those described in the present embodiment, and Equivalent techniques are also included in the scope of the present invention. The present invention can be applied to various games. Further, the present invention is applied to various image generation systems such as a business game system, a home game system, a large attraction system in which a large number of players participate, a simulator, a multimedia terminal, a system board for generating a game image, and a mobile phone. it can.

The example of the block diagram of the image generation system of this embodiment. The figure explaining an example of the stereoscopic vision system which this embodiment employ | adopts. The figure explaining an example of the stereoscopic vision system which this embodiment employ | adopts. 4A and 4B are examples of the right-eye image and the left-eye image generated according to this embodiment. The example of the multi-view stereoscopic image produced | generated by this embodiment. Explanatory drawing of the multi-view stereoscopic image generation method of this embodiment. FIGS. 7A and 7B are explanatory diagrams of copy processing of the right-eye image and the left-eye image. FIGS. 8A and 8B are explanatory diagrams of copy processing using texture mapping. FIG. 9A and FIG. 9B are explanatory diagrams of the method of this embodiment when the double buffer method is adopted. Explanatory drawing of the averaging process method of this embodiment. Explanatory drawing of the averaging process method of this embodiment. An example of an image generated when the averaging method of this embodiment is adopted. The example of the image produced | generated when the averaging method of this embodiment is not employ | adopted. Explanatory drawing of the method of shifting the gaze point of a virtual camera. The flowchart explaining the detailed process of this embodiment. The flowchart explaining the detailed process of this embodiment. FIG. 17A and FIG. 17B are hardware configuration examples.

Explanation of symbols

DB drawing buffer, DB1, DB2 first and second drawing buffers,
TB buffer, AR1, AR2 first and second areas, MS model data storage unit,
TS texture storage unit, MTS motion data storage unit,
100 processing unit, 102 game calculation unit, 104 object space setting unit,
105 culling processing unit, 106 moving object calculation unit, 108 virtual camera control unit,
120 drawing units, 122 lighting processing units, 124 texture mapping units,
130 sound generation unit, 160 operation unit, 170 storage unit, 180 information storage medium,
190 Display unit, 192 Sound output unit, 194 Auxiliary storage device, 196 Communication unit

Claims (14)

A buffer having a first region and a second region;
As a drawing unit that performs drawing processing to generate a multi-view stereoscopic image,
Make the computer work,
The drawing unit
The first viewpoint image for multi-view stereoscopic viewing is compressed in half in the height direction and written to the first area of the buffer, and the second viewpoint image for multi-view stereoscopic viewing is viewed in the height direction. Compressed to 1/2 and written to the second area of the buffer,
By reading the image from the buffer, the pixel image of the first viewpoint image written in the first area and the pixel image of the second viewpoint image written in the second area A program for generating the multi-view stereoscopic image displayed alternately every time. In claim 1,
The buffer is
The first region having a width of W pixels and a height of H / 2 pixels; and the second region having a width of W pixels and a height of H / 2 pixels;
The drawing unit
The memory pitch of the buffer is set to 2N bytes, the first viewpoint image having a width of W pixels and a height of H pixels is compressed in the height direction, written to the first area, and the width Compresses the second viewpoint image in which W is W pixels and H is H pixels in the height direction, and writes the compressed image to the second area,
By setting the memory pitch of the buffer to N bytes and reading the image from the buffer, the pixel image of the first viewpoint image and the pixel image of the second viewpoint image are alternately displayed for each line. A program for generating a multi-view stereoscopic image. In claim 1 or 2,
A drawing buffer for drawing the first and second viewpoint images;
The drawing unit
The first viewpoint image is drawn in the drawing buffer, the first viewpoint image drawn in the drawing buffer is compressed in the height direction, copied to the first area of the buffer, and the second The viewpoint image is drawn in the drawing buffer, and the second viewpoint image drawn in the drawing buffer is compressed in the height direction and copied to the second area of the buffer. In claim 3,
The drawing unit
The first viewpoint image drawn in the drawing buffer is texture-mapped to a polygon having the size of the first area and is drawn in the first area, so that the drawing buffer transfers to the first area. The first viewpoint image is copied, and the second viewpoint image drawn in the drawing buffer is texture-mapped to a polygon having the size of the second area and drawn in the second area. The program for copying the second viewpoint image from the drawing buffer to the second area. In claim 4,
The drawing unit
A program for shifting texture coordinates in the height direction by one texel when texture mapping the second viewpoint image onto a polygon having the size of the second region. In any of claims 3 to 5,
The drawing unit
A program for writing back the buffer image to the drawing buffer after the first and second viewpoint images are copied to the first and second areas. In claim 6,
The drawing unit
A program characterized in that an image in the buffer is texture-mapped to a polygon of the size of the drawing buffer and drawn in the drawing buffer, whereby the image in the buffer is written back into the drawing buffer. In claim 6 or 7,
As the drawing buffer, first and second drawing buffers are prepared,
The drawing unit
After performing image copy processing from one drawing buffer of the first and second drawing buffers to the buffer and image writing back processing from the buffer to the one drawing buffer, the one drawing buffer A program that switches from a back buffer to a front buffer and switches the other drawing buffer different from the one from the front buffer to the back buffer. In any one of Claims 1 thru | or 8.
The drawing unit
When the first and second viewpoint images are compressed in the height direction and written in the first and second areas, the pixel image of the i-th line of the first and second viewpoint images and the next A program for averaging pixel images of the (i + 1) th line. In claim 9,
The drawing unit
The first and second viewpoint images are texture-mapped to the polygons having the sizes of the first and second regions by a texel interpolation method, and are drawn in the first and second regions. A program for performing the averaging process on a pixel image of a line and a pixel image of the (i + 1) -th line. In claim 9 or 10,
The drawing unit
For the m-th line of the multi-view stereoscopic image, the pixel image of the j-th line of the first viewpoint image and the pixel image of the next j + 1-th line are averaged, and the multi-view stereoscopic view is obtained. For the m + 1st line of the image, the averaging process is performed so that the pixel image of the j + 1th line of the second viewpoint image and the pixel image of the next j + 2 line are averaged. A featured program. In any one of Claims 1 thru | or 11,
The drawing unit
The gazing point of the virtual camera for the first viewpoint of multi-view stereoscopic vision and the gazing point of the virtual camera for the second viewpoint of multi-view stereoscopic viewing are shifted in the height direction, and the first viewpoint image and A program for generating the second viewpoint image.   A computer-readable information storage medium, wherein the program according to any one of claims 1 to 12 is stored. An image generation system for multi-view stereoscopic viewing,
A buffer having a first region and a second region;
A drawing unit that performs a drawing process for generating a multi-view stereoscopic image;
Including
The drawing unit
The first viewpoint image for multi-view stereoscopic viewing is compressed in half in the height direction and written to the first area of the buffer, and the second viewpoint image for multi-view stereoscopic viewing is viewed in the height direction. Compressed to 1/2 and written to the second area of the buffer,
By reading the image from the buffer, the pixel image of the first viewpoint image written in the first area and the pixel image of the second viewpoint image written in the second area An image generation system for generating the multi-view stereoscopic image displayed alternately every time.