JP2000030042A - Display processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the hardware operation efficiency and to reduce the hardware scale by blending pixel values of reference frame data and difference frame data written in a 1st and a 2nd memory at a specific mixing ratio. SOLUTION: A compositing part 9 composites the frame image data stored in a frame memory 2 and rendering data after the pasting of a bit map and color painting by a texture mapping part 8. This compositing is carried out by calculating the transmissivity of each body according to the transparency set to the body and illuminance set for the vertexes of the body and mixing pixels having common coordinates between pixels constituting the frame image data and pixels constituting the rendering data according to the transmissivity of each body.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a display processing device for performing display processing on moving image data and three-dimensional graphics data.

[0002]

2. Description of the Related Art Multimedia works using CD-ROMs and DVD-ROMs as recording media have been published one after another in service industries such as entertainment, music, and education. There are various types of what are recognized as multimedia works in the business world. Among those types, those which are particularly expected to develop in the future are multimedia works mainly composed of three-dimensional graphics data and multimedia works composed of moving image data.

[0003] The three-dimensional graphics data includes a shape,
It consists of data that quantifies the physical properties of an object, such as spatial arrangement and material, and the playback device uses these data to reproduce the physical properties of the object and display an image that depicts it. Means that the object appears as if it were in front of the operator. Multimedia works using three-dimensional graphics data include drive games and flight simulation games full of speed, virtual walk-through systems that give operators the feeling of walking in virtual stores and virtual exhibition halls. is there.

[0004] The moving image data is data obtained by compressing and encoding an analog video signal using an intra-frame encoding technique and an inter-frame encoding technique, and includes a plurality of frame data. Multimedia works using this include theatrical movies mainly composed of live-action footage,
Document videos and TV dramas are famous. There are the above two types of multimedia works, which are gaining much support from consumers and traders.
There is a growing demand for a playback device that can play both types of multimedia works. In this case, the playback device must implement a playback function of three-dimensional graphics data by mounting a dedicated board having a computing capability equal to the hardware mounted on a home TV game machine or the like. MPEG implemented in DVD players, etc.
A dedicated board having a decoding capability equal to that of a decoder must be mounted to realize a function of reproducing moving image data.

[0005]

However, the above-mentioned 2)
A playback device having a playback function for two types of multimedia works requires a dedicated board for each of three-dimensional graphics data and moving image data, even if the playback function is to be realized. The area occupied by the two dedicated boards and the manufacturing cost weigh heavily on the production cost of the playback device, and the size of the playback device is reduced.
There is a problem that the cost reduction is greatly hindered.

Also, it is rare that the reproduction of a multimedia work based on a three-dimensional graphics data system and the reproduction of a multimedia work based on a moving image data system are rarely performed simultaneously. While one of the multimedia works of the image data system is being played, one of the two dedicated boards is wasting time without performing any processing, and the operation of the hardware is performed. There is a problem that the efficiency is extremely low.

[0007] It is an object of the present invention to provide a display processing device capable of reducing the hardware scale and improving the cost of the reproduction device by improving the hardware operation efficiency.

[0008]

In order to achieve the above-mentioned object, the present invention provides a method for rendering a plurality of rendering data indicating a plurality of projection images to be combined and displayed, and each of the overlapping rendering data at the time of the combination. A display processing device that performs display processing on three-dimensional graphics data including transparency information indicating how much light is transmitted, and performs display processing on moving image data including a plurality of frame data. A first memory, a second memory, a receiving unit that receives an instruction from the operator for one of the display process for the image data and the display process for the moving image data,
A first writing unit that, when receiving a display processing instruction for the three-dimensional graphics data, writes the rendering data of the plurality of rendering data to a first memory and writes the rendering data of a plurality of rendering data to a second memory; Blending means for blending a pixel value of the background-side rendering data written to the first memory and a pixel value of the foreground-side rendering data written to the second memory at a mixing ratio based on the transparency information; When an instruction for data display processing is received, difference frame data including only difference information indicating a difference from previous and subsequent frame data among a plurality of frame data, and reference frame data including only difference value data and a pixel value component to be added are included. And a second writing means for writing the data into the first memory and the second memory, respectively. When the difference frame data and the reference frame data are written to the first memory and the second memory, the row includes a pixel value of the reference frame data written to the first memory and the second memory, and a pixel value of the difference frame data. Are blended at a predetermined mixing ratio.

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, as a preferred embodiment of a display processing apparatus according to the present invention, a multimedia reproducing apparatus having a display processing apparatus therein will be described. FIG. 1 is a diagram showing an AV system having a multimedia playback device as a core. In this figure, a multimedia reproducing apparatus 300 is an optical disc 30 that is a recording medium of a multimedia work.
1. The optical disc 302 is loaded, the multimedia work recorded on the optical disc 302 is reproduced, and the reproduced video is displayed on the television receiver 303. Here, the optical disk 30
Numeral 1 is for recording three-dimensional graphics data, and optical disc 302 is for recording moving image data. Prior to the description of the multimedia reproducing apparatus, the data structures of three-dimensional graphics data recorded on the optical disk 301 and moving image data recorded on the optical disk 302 will be described.

FIG. 2 is a diagram showing a data structure of three-dimensional graphics data. As shown in the figure, the three-dimensional graphics data includes background image data k1 and light source coordinates k2.
And luminous intensity k3, object information k4 about a plurality of objects to be arranged in the virtual space, a viewpoint position k5, and a line-of-sight direction k6. With only the above data recorded, the optical disk 3
01 is carried.

The background image data k1 is image data m1 which is an image for one frame, and the light source coordinates k2 are virtual space coordinates including an X-axis component, a Y-axis component, and a Z-axis component. The object information k4 includes the object name g1, the arrangement coordinates g2 where the object is to be arranged in the virtual space, the shape information g3 representing the shape specific to the object in a relative coordinate sequence, and a pattern attached to the object plane. Pattern information g4 indicating the bitmap to be attached
And color information g5 indicating the color to be colored on the object, reflectivity g6 indicating how much the material of the object reflects the light emitted from the light source, and how much light the object emits from the light source And transparency g7 indicating whether the light is transmitted.

FIG. 3 shows the shape of each relative coordinate sequence specified in each object information. The relative coordinate sequence of the object “ice pillar” represents a cubic shape as shown in FIG. 3A, and the relative coordinate sequence of the object “Piramit” represents a triangular pyramid shape as shown in FIG. 3B. .
The relative coordinate sequence of the object "meteorite" represents a spherical shape as shown in FIG.

FIG. 4 shows what kind of pattern each pattern information specified in each object information represents. The pattern information of the object "ice pillar" represents the spelling "ICE" of the alphabet as shown in FIG. 4A, and the pattern information of the object "Piramit" represents a stone wall as shown in FIG. 4B. . The pattern information of the object "meteorite" represents a rock surface as shown in FIG. These are for giving a material feeling or the like to a substance expressed only by vertices.

Here, the arrangement coordinates (x1, y1, z
1) to (x3, y3, z3), light source coordinates (xn, yn, zn), and viewpoint position (x
0, y0, z0) and the line-of-sight vector (Vx, Vy, Vz) are shown in FIG. FIG. 5 is a diagram showing how the respective coordinates in the three-dimensional graphics data are in a positional relationship. In this case, assuming that the operator looks at the scene where these objects are arranged in the direction of the line of sight from the viewpoint position (x0, y0, z0), the operator can see the object icicles in the foreground, and behind them the pyramid and meteorite You should see the background image. In consideration of such an arrangement relationship, the multimedia playback device
Generate a projection image to appear on the screen m2 in the figure as a view from the viewpoint. FIG. 6 shows a case where the coordinates in the three-dimensional graphics data have the positional relationship shown in FIG.
FIG. 14 is a view showing a projected image to be displayed on a screen m5.

Next, the data structure of moving image data recorded on the optical disk 302 will be described. Optical disk 3
The moving image data recorded in No. 02 has a data structure shown in FIGS. The moving image data is composed of a huge number of GOP data of hundreds or thousands. Here, the GOP data is a group of frame data to be displayed in a video display period of 0.5 seconds. FIG.
The first GOP data in (a) is a group of frame data to be displayed from the display time 00.0 seconds to 00.5 seconds, and the second and third GOP data are displayed from the display time 01.5 seconds to 02.0 seconds. This is a collection of frame data to be displayed, and frame data to be displayed from a display time of 02.5 seconds to 03.0 seconds. The moving image data recorded on the optical disc 302 is composed of a series of GOP data.
It constitutes a moving image such as a drama.

The frame data included in each GOP data is 12 frames to 1 to be displayed in the time unit of 0.5 seconds.
It corresponds to video frames such as five. FIG. 7B shows the data structure of the frame data. Each frame data includes display order information k7 indicating its own display order, among I pictures, B pictures, and P pictures specified by the MPEG standard.
Picture type k8 indicating to which type its own belongs,
In the case of a P picture or a B picture, it indicates which macroblock is located in itself, and motion vector information k9 indicating which direction the macroblock is located has a high correlation with the macroblock in which it is located, and includes a plurality of pixel values. Slice data k10
Consists of

An I-picture is a compressed image data that includes a luminance component and a color difference component for one frame of pixels. P picture (Predictive-Pict
ure) or B picture (Bidirectionally predictive Pic)
ture) does not have a pixel value for one frame but includes a pixel value (called a difference image) of a portion that has changed compared to the previous and next frames. Here, the P picture is a difference image including a difference from a frame located in the past direction, and the B picture is a difference image including a difference from a frame located in the past direction and the future direction.

Here, how the slice data included in the I picture, the P picture, and the B picture are different will be described with reference to FIG. FIG. 8 shows the magnitude of the slice data as I
FIG. 9 is an explanatory diagram for comparing a picture, a P picture, and a B picture. In FIG. 8, a hatched portion indicates a portion where the macroblock does not exist. The slice data for an I picture is composed of 45 macroblocks.

A macroblock is a luminance block composed of 16 × 16 luminance components, a blue difference block (Cb block) composed of 8 × 8 blue difference components, and a 8 × 8 red color difference block. This is data including a red difference block (Cr block) composed of components, and pixel values in moving image data are written to a memory in units of macroblocks and read from the memory.

FIG. 8 also shows that the I picture has pixel values of all macroblocks since there is no hatched portion in the I picture. On the other hand, it can be seen that most of the P picture and the B picture are hatched, and do not have the pixel value of the macroblock. The slice data of the P picture and the B picture have macroblocks only in the macroblocks related to the part where the bird is flying in the I picture.

The reason why the P picture and the B picture have macroblocks only in the portions where the video contents have changed as compared with the frame data before and after the P and B pictures is to compress the data amount. In the decoding process (the decoding process here is the same as the display process in the claims), the compressed moving image data is used as a difference image and an image serving as a reference for calculating the difference (hereinafter referred to as a “reference image”). ) Is converted into an image for display.

Although it is clear from the above description that the three-dimensional graphics data and the moving image data have significantly different data structures and information contents, the display processing of the three-dimensional graphics data and the decoding processing of the moving image data This embodiment focuses on the existence of the following common points (1) and (2). <Common point (1)> At the time of reproducing three-dimensional graphics data, if a plurality of objects are arranged in a virtual space, a projection image must be generated for each of these objects. Here, of the objects having transparency, a projection image on the front side and a projection image on the back side are respectively generated. The images to be displayed next must be obtained by combining the projection images generated in this way. When the projection image of the example of FIG. 6 is generated, since the material of the icicle has transparency, the multimedia reproducing apparatus performs four projection images, that is, a projection image on the front side of the icicle and a projection image on the back side of the icicle. A projection image, a projection image on the front side of the pyramid, and the like are separately generated, and these are combined as shown in FIG. 9 to generate a scene image of a virtual space that can be seen from the viewpoint.

On the other hand, when the next frame data is a P-picture at the time of reproducing the moving picture data, the P-picture is decoded to obtain the difference image, and the reference picture to be added to the difference picture from the I-picture is obtained. , And the difference image and the reference image are combined as shown in FIG. 10 to obtain a display image of the next frame. Moving image data and 3
Although the two-dimensional graphics data has a completely different data structure, they have in common that a plurality of images need to be combined. If the hardware is configured with this commonality in mind, a memory (generally called a texture memory) for individually storing a plurality of images to be combined and the pixel values of the plurality of images to be combined are added. The adder to be combined can be shared.

<Common point (2)> At the time of decoding moving image data, if the next frame data is a P picture or a B picture, the cut-out range of the reference image indicated by the motion vector information is a half pixel ahead. There may be. In order to interpolate the displacement in this case, an interpolation process called half-pel interpolation is required.

At the time of display processing of the three-dimensional graphics data, when generating a projection image of the object arranged in the virtual space on the screen m2, the projection image is arranged in the line-of-sight direction from the vertex coordinates and the viewpoint coordinates of the object in the virtual space. It is necessary to calculate the projection point coordinates with the screen m2. When the projection point coordinates are not integer values, an interpolation process called half-pel interpolation is required as in the case of moving image data.

Considering the above commonality, it seems that the adder for the combining process and the dedicated circuit for the half-pel interpolation can be simply shared. However, since the adder for the synthesis processing has the following differences (1) and (2), the display processing of the three-dimensional graphics data and the moving image data does not simply use the common processing. . <Difference (1)> Since image data is a set of pixel values, a negative value cannot be originally provided. The same applies to pixel values obtained by performing display processing on three-dimensional graphics data, and all have positive pixel value components. On the other hand, since the difference image included in the B picture and the P picture is decoded based on the difference from the preceding and succeeding frame data, and has a negative component, the processing of synthesizing the three-dimensional graphics data is performed. Cannot process pixel values having such a negative component. <Difference (2)> In the display processing of three-dimensional graphics data, a front projected image of a transparent object and a rear projected image are displayed. When synthesizing a projection image, it is necessary to mix pixel values of these two projection images at a ratio corresponding to the transparency, and such a function is required for an adder. Does not require such a function.

Considering the difference (2) among the above differences, the adder provided for the display processing of the three-dimensional graphics data has a function of mixing pixel values at a predetermined ratio. Can be said to be excellent. However, only the adder for 3D graphics data is different
Signed arithmetic processing required for (1) cannot be realized. For this reason, the common use of the adder for image synthesis in the present embodiment is based on an adder for three-dimensional graphics data display processing having a function of mixing pixel values, which is used to realize signed arithmetic processing. It is considered wise to add various functions. When decoding moving image data,
The setting of the transmittance in the adder for three-dimensional graphics data will be described later.

<Internal Configuration of Multimedia Reproducing Apparatus> FIG. 11 is a diagram showing the inside of a multimedia reproducing apparatus. In FIG. 11, the multimedia playback device 300
A drive device 101 for loading a disk-type recording medium such as an optical disk 301 and an optical disk 302 to read out three-dimensional graphics data and moving image data, a display processing device 102 according to the present embodiment, and a display processing device 1
The video signal conversion device 103 converts the display processing result of the image processing device 02 into a video signal. Display processing device 1
FIG. 12 shows the internal configuration of the H.02.

In FIG. 12, the display processing device 102 includes an input switching unit 1 and a three-dimensional graphics data dependent unit 20.
0, the moving image data dependent unit 201, the memory module 202, the synthesizing unit 9, the reading address generating unit 14, the bidirectional half-pel interpolating units 19a, b, c, d, e. It comprises a unit 20 and a switch 37. Among them, the graphic data dependent unit 200 and the moving image data dependent unit 20
Numeral 1 depends on three-dimensional graphics data display processing and moving image data display processing, respectively, and other parts (memory module 202, synthesizing section 9, reading address generating section 14, bidirectional half-pel interpolation sections 19a, b, c) , d, e..., the sign bit determination unit 20 and the switch 27) are shared (1),
It has been standardized focusing on (2).

The input switching section 1 refers to the index information of the recording medium loaded in the decoding control section 101 and determines whether the data recorded on the recording medium is three-dimensional graphics data or moving image data. It is determined whether there is. If the index information indicates three-dimensional graphics data, the data read from the drive device 101 is output to the graphic data dependent unit 200 as three-dimensional graphics data. If moving image data is indicated in the index information, the data read from the drive device 101 is output to the moving image data dependent unit 201 as moving image data.

The memory module 202 is a memory board including a frame memory 2, a texture memory array 3, and a work memory 4. Frame memory 2
Is a memory corresponding to the screen m2 arranged in the virtual space, and the pixel values stored therein are sequentially converted into video signals by the video signal conversion device 103.

When displaying three-dimensional graphics data, the texture memory array 3 is provided with a plurality of texture memories for storing rendering data on the front side and rendering data on the back side of an object corresponding to each object information. Become. Here, the rendering data refers to data that expresses a projected image of an object arranged in the virtual space for each of the front side and the back side.

The first object front side texture memory 31 of the texture memory array 3 stores front side rendering data of an object located at the foremost side (referred to as a first object) when viewed from the viewpoint. The object back side texture memory 32 stores the back side rendering data of the first object located closest to the viewer when viewed from the viewpoint. The second object front side texture memory 33 stores front side rendering data of an object (referred to as a second object) located behind the first object when viewed from the viewpoint, and the second object back side texture memory 34 stores , Which stores the rendering data on the back side of the second object located behind the first object when viewed from the viewpoint. The third object front side texture memory 35 stores front side rendering data of an object (referred to as a third object) located behind the second object when viewed from the viewpoint, and the third object back side texture memory 36 stores Rendering data on the back side of a third object located behind the second object when viewed from the viewpoint. As described above, one texture memory is allocated to each of the front side rendering data and the back side rendering data of the first, second, and third objects.
It can be seen that the front side rendering data and the back side rendering data of one object are stored separately.

The work memory 4 is used for generating rendering data when displaying three-dimensional graphics data. Also, it is used for decoding an I picture or a difference image at the time of decoding moving image data. Note that the capacities of the frame memory 2, texture memory array 3, and work memory 4 described above each have a capacity to hold a pixel value for one screen, and are connected to a data bus and an address bus.

The internal areas of the frame memory 2, texture memory array 3, and work memory 4 are designated by a common absolute address. The absolute address is composed of a base address and an offset address, and the head address assigned to each of the frame memory 2, the texture memory array 3, and the work memory 4 is used as the base address. The offset address is used to address a storage area of each pixel value located at an arbitrary coordinate (x, y) on the image in the frame memory 2, the texture memory array 3, and the work memory 4 (the offset address is Since the offset address is used exclusively for designating the storage destination of the pixel value, the offset address is hereinafter treated equivalently to the XY coordinates.)

Further, a frame memory 2 to a work memory 4
Holds the pixel values of the three-dimensional graphics data based on the values of the luminance components for each of the three primary colors of red, blue, and green. For moving image data, the values of the luminance component, the blue difference component, and the red difference component Represents the pixel value. Next, the internal configuration of the graphic data dependent unit 200 will be described. The graphic data dependent unit 200
Geometry unit 5, inclination calculation unit 6, shading unit 7,
It is constituted by a texture mapping unit 8.

The geometry section 5 detects a viewpoint position and light source coordinates included in the three-dimensional graphics data,
The viewpoint position and the light source coordinates are determined for these coordinates. Also, the object vertices are arranged in the virtual space by reading the object information included in the three-dimensional graphics data and adding the relative coordinate sequence included in the shape information to the arrangement position coordinates included therein. . Finally, the vector indicating the line-of-sight direction is defined as its normal vector, and a plane having spatial coordinates on an extension line from the viewpoint coordinates toward the line-of-sight vector is calculated. Set to.

The inclination calculator 6 sets each side of the object in the virtual space as a straight line connecting the vertices, and calculates the inclination formed by the straight line between the sides and the screen. After the calculation, the coordinates of the projection point of the object vertex are obtained on the screen set in the virtual space, and the projection point coordinates are grouped for each of the planes and written in the work memory 4.
Get the rendering data prototype. At this time, non-integer coordinates among projection point coordinates constituting rendering data written in the work memory 4 are rounded to integer values.

Here, the icicle and pyramid are in the geometry section 5
When placed in the posture shown in FIG.
Since the projection point coordinates of the vertices constituting the icicle and pyramid are obtained, the prototypes of the rendering data shown in FIGS. 20 (a) to (f) and FIGS. 21 (a) and (b) are obtained. FIG. 20 and FIG.
FIG. 1 is a diagram showing what rendering data is obtained by the icicle and pyramid in the posture of FIG.

The shading unit 7 calculates the illuminance of each point of the object in the virtual space. The illuminance to be calculated here is distributed to each object from the light source of luminous intensity R1 arranged at the light source coordinates (xn, yn, zn), the incident angle of the light ray on the surface including each vertex, the light source It is calculated from the distance between the point and the vertex, and the reflectance of the surface. The illuminance is a value from 0 to 1 and if the illuminance is 1,
The color information in the object information is displayed as it is,
If so, the color information is lost.

The texture mapping unit 8 integrates, among the rendering data grouped for each object plane in the work memory 4, those located on the front side and the back side as viewed from the viewpoint, and then combines these rendering data. A bitmap is pasted on a closed area encompassed by the relative coordinate sequence in and color is filled with the designated color information. After the filling, the filled rendering data is shaded based on the illuminance for each vertex calculated by the shading unit 7. When the shadow is added in this way, the rendering data is written to the corresponding texture memory in the texture memory array 3.

When the above processing is performed on the rendering data shown in the examples of FIGS. 20 and 21, the rendering data of the icicle closest to the viewpoint is used as the first object to integrate the planes. The front side in FIG. 20 is shown in FIG.
FIG. 2 shows a plane composed of vertices P3, P4, P5, and P6 shown in FIG.
This is a plane including vertices P2, P4, P6, and P7 shown in FIG.
The texture mapping unit 8 integrates these into one rendering data to generate the front side rendering data of the first object. Thereafter, based on the pattern information specified for the icicle, the character string ICE is pasted to obtain the rendering data shown in FIG. 22A, and the first object assigned to the front-side rendering data of the first object Write to the front side texture memory 31.

FIG. 20 shows the back side of the first object.
A plane composed of vertices P1, P2, P3, and P4 shown in FIG.
FIG. 20 is a plane including vertices P5, P6, P7, and P8 shown in FIG.
FIG. 20F shows a plane including P1, P3, P5, and P8 shown in FIG.
Is a plane including P1, P2, P7, and P8 shown in FIG. The shading unit 7 integrates these into one rendering data to generate the rendering data on the back side of the first object. Then, based on the pattern information specified for the icicle,
The rendering data shown in FIG. 22B is obtained by pasting the ICE, and is written in the first object back side texture memory 32 assigned to the back side rendering data.

Next, the rendering data of the pyramid closest to the viewpoint is used as the second object to integrate the planes. The front side of the second object is a plane formed by vertices Q1, Q2, and Q3 shown in FIG. 21A and a plane formed by vertices Q1, Q3, and Q4 shown in FIG. 21B. Shading part 7
Integrates these into one rendering data to generate front-side rendering data of the second object. Then, based on the pattern information specified in the pyramid, a stone wall pattern is pasted to obtain rendering data shown in FIG. 22C, and the second object front surface assigned to the front side rendering data of the second object is obtained. Writing to the side texture memory 33.

The synthesizing unit 9 synthesizes the frame image data stored in the frame memory 2 and the rendering data to which the bitmap is pasted by the texture mapping unit 8 and the color is filled. The synthesis here calculates the transmittance of each object based on the transparency set for each object and the illuminance set for the vertices of each object, and configures the pixels constituting the frame image data and the rendering data. This is performed by mixing the pixels having the same coordinates on the image among the pixels to be processed according to the transmittance of each object.

Here, the transmittance to be calculated is as follows: transmittance = 100% at a point on a transparent surface, transmittance = 0% at a point on an opaque surface, and At some point it is greater than 0 and less than 100%. The significance of performing the mixing calculation based on such transmittance will be described. Here, it is assumed that the front-side projection image of the object A located closest to the viewpoint with respect to the viewpoint has pixel values (Rp, Bp, Gp).
Also, it is assumed that the back side projection image of the object B located on the back side of the object A has a pixel value (Rr, Br, Gr). Then, the pixel values (Rp,
Bp, Gp) should appear at the rate of α%, and the pixel values (Rr, Br, Gr) corresponding to the projected image of the object B should appear on the screen at the rate of (100−α)%.

The pixel value (Rr, Br, Gr) shown by the following equation appears on the screen due to the object A having the transmittance α. Rr = Rp × α + Rq × (100-α) Br = Bp × α + Bq × (100-α) Gr = Gp × α + Gq × (100-α) Since this is equivalent to the operation of adding two pixel values, the synthesizing unit 9 performs mixing of the pixel values by performing the operation of the above formula based on the pixel value of each object and the transmittance.

This concludes the description of the internal configuration of graphic data dependent unit 200. Next, the internal configuration of the moving image data dependent unit 201 will be described. As shown in FIG. 12, the moving image data dependent unit 201
, A variable-length decoding unit 11, an inverse quantization unit 12, and an inverse DCT unit 13. The decoding control unit 10 divides the moving image data into GOP data, and divides the data into picture data and further into macroblocks. After completion of the decoding process, each macroblock is added with management information such as identification information of the stream data and picture to which it belongs so that it can be reconstructed into a picture or GOP, and then transmitted to the variable length decoding unit 11.

The variable length decoding unit 11 performs a variable length decoding process on the data of the transmitted macro block,
The quantization coefficient of one macroblock (three-layer structure of 4 + 1 + 1 blocks) is obtained, and the position information (position in the picture) of the macroblock and the motion vector are extracted from the management information and output to the inverse quantization unit 12. . Inverse quantization unit 1
2 dequantizes the quantized coefficient of one macroblock output from the variable length decoding unit 11 to obtain a DCT coefficient,
It is sent to the inverse DCT section 13.

The inverse DCT unit 13 performs an inverse frequency transform (inverse DCT) on the DCT coefficient transmitted from the inverse quantization unit 12.
T) Perform a process to obtain a pixel value and write it to the work memory 4. This concludes the description of the components of the moving image data dependent unit 201 provided exclusively for moving image data. Hereinafter, the components commonly used in the moving image data decoding processing and the three-dimensional graphics data display processing will be described.

The reading address generator 14 stores the first to n-th rendering data to which the bit map is pasted and colored in the work memory 4 by the texture mapping unit 8, or vice versa. DCT
When storing the processing result by the unit 13 in the work memory 4,
A write destination address having the work memory 4 as a write destination is generated and output to an address bus. Further, when transmitting the data stored in the work memory 4 to the texture memory array 3 and the frame memory 2, a write destination address to be a write destination is generated based on the content of the data to be transmitted, and the data is transmitted to the address bus. Output. In order to perform the above processing, the reading address generator 14 has an internal configuration shown in FIG. FIG. 13 is a diagram showing the internal configuration of the reading address generation unit 14. As shown in FIG. 13, the base address generation unit 15, the rendering data offset address generation unit 16, the difference image offset address generation unit 17, and the reference image The offset address generation unit 18 is provided.

The base address generation unit 15 includes a frame memory 2, a texture memory array 3, and a work memory 4.
At the time of storing data or transmitting data between the work memory 4, the frame memory 2 and the work memory 4,
A base address of a read destination and a write destination is generated and output to an address bus. When storing the front-side and back-side rendering data of the first to n-th objects in the work memory 4, or vice versa
When storing the processing result of the DCT unit 13 in the work memory 4, the head address of the work memory 4 is generated as a base address and output to the address bus. When transmitting the data stored in the work memory 4 to the texture memory array 3 and the frame memory 2, the data is generated as a base address of a write destination based on the correspondence table shown in FIG.

As shown in FIG. 14, if the data to be transmitted is rendering data,
The head address of the work memory 4 is generated as a read destination base address and output to the address bus, and the head address of each texture memory in the texture memory array 3 is generated as a write destination base address and output to the address bus.

If the data to be transmitted is a differential image, the base address generator 15 generates the head address of the work memory 4 as a read destination base address and outputs the generated base address to the address bus. The head address of the texture memory 31 is generated as a write destination base address and output to the address bus.

If the data to be transmitted is a reference image, the base address generator 15 generates the head address of the frame memory 2 as the base address of the read destination and outputs the generated base address to the address bus. The head addresses of the texture memory 33 and the third object front side texture memory 35 are generated as base addresses for writing, and output to the address bus.

If the data to be transmitted is an I picture, the head address of the work memory 4 is generated as the base address of the read destination address and output to the address bus, and the head address of the frame memory 2 is generated as the base address of the write destination. And outputs it to the address bus. As described above, when the base address generation unit 15 generates the base address and outputs the generated base address to the address bus, the rendering data, the reference image, the difference image, and the I picture are transferred to the appropriate memories.

When the rendering data, the reference image, and the difference image are transmitted to the texture memory array 3 based on the correspondence table in FIG. 14, the rendering data, the reference image, and the difference image are transmitted as shown in FIGS. The images are stored in the first object front side texture memory 31 to the third object back side texture memory 35. The three rendering data shown in FIG. 22 are generated in the work memory 4, and at the time of generation of each rendering data, the base address generating unit 15
When the head addresses of the first to third object back side texture memories 35 are generated as base addresses and output to the address bus, the first object front side texture memories 31 to the third object back side textures are generated as shown in FIG. Memory 3
5 stores the rendering data shown in FIGS.

Also, at the time of decoding the P picture, the base address generation unit 15 generates the head address of the first object front side texture memory 31 as the base address of the writing destination of the difference image, and generates the second object front side texture memory. When the head address of the reference image 33 is output as the base address of the writing destination of the reference image, the difference image is stored in the first object front side texture memory 31 as shown in FIG.
The reference image is stored in the second object front side texture memory 33.

Further, at the time of decoding the B picture, the base address generator 15 generates the head address of the first object front side texture memory 31 as the base address of the difference image writing destination, and generates the second object front side texture memory. 33, when the head address of the third object front side texture memory 35 is output as the base address of the writing destination of the front side reference image and the rear side reference image, as shown in FIG. The difference image is stored in 31, and the front-side reference image and the rear-side reference image are stored in the second object front side texture memory 33 and the third object front side texture memory 35.

The storing of the difference image in the first object front side texture memory 31 is merely an example for convenience of explanation. The difference image and the reference image are stored in the first object front side texture memory 31 to the third object back side texture memory 36.
Which of the above is stored may be arbitrarily determined. The rendering data offset address generator 16 generates an offset address in the work memory 4 based on the XY coordinates calculated by the tilt calculator 6, and outputs the offset address to the address bus.

The difference image offset address generator 17
Is that the picture type in the newly decoded frame data is B picture or P picture, and when the difference image is stored in the work memory 4, the macro block address extracted from the header by the inverse quantization unit 12 is set to the xy coordinate. By performing the conversion, the offset address of the write destination is generated and output to the address bus. After the output of the offset address, when the inverse DCT unit 13 outputs the pixel value of the difference image,
The pixel value is taken into the work memory 4 specified as the write destination address, and written into the internal area indicated by the offset address inside the work memory 4.

Reference image offset address generator 18
In the newly decoded frame data, the picture types are B picture and P picture, and when the reference image is cut out to the frame memory 2, the macroblock address and the motion vector information extracted from the header by the inverse quantization unit 12 are Is converted to xy coordinates to generate a read destination offset address and output it to the address bus. After the output of the read destination address, the frame address is fetched into the frame memory 2 designated as the read destination address, and
Data is read from the internal area indicated by the internal offset address and output to the data bus.

Next, the components interposed between the texture memory array 3 and the data bus will be described. FIG. 16 is a diagram showing components of the texture memory array 3 interposed between the first object front side texture memory 31 and the data bus. The bidirectional half-pel interpolation unit 19a is provided in association with the first object front side texture memory 31. When the half-pel interpolation is required, the bidirectional half-pel interpolation unit 19a is used for the pixel value read from the first object front side texture memory 31. Performs bidirectional and unidirectional half-pel interpolation.

In the display processing of moving image data, half-pel interpolation is required when the extraction range of the reference image specified by the motion vector information is a half pixel ahead, and the display processing of three-dimensional graphics data is performed. In the above, half-pel interpolation is required when the projection point coordinates of the screen m2 arranged in the line of sight from the vertex coordinates and the viewpoint coordinates of the object in the virtual space are not integer values.

If one-way half-pel interpolation is necessary, the pixel value of the coordinate (x + 1, y) or the coordinate (x + 1, y) is read out together with the pixel value of the coordinate (x, y) read from the first object front side texture memory 31.
The pixel value of (x, y + 1) is read, the average value of these pixel values is calculated, and the pixel value of coordinates (x, y) is replaced with the calculated average value and output to the data bus. When bi-directional half-pel interpolation is required, the pixel value of the coordinates (x, y) and the pixel value of the coordinates (x + 1, y) and the coordinates (x, y) are read out together with the pixel value of the coordinates (x, y) read from the texture memory of the texture memory array 3. y + 1) and the pixel value at coordinates (x + 1, y + 1) are read, the average of these pixel values is calculated, and the pixel value at coordinates (x, y) is calculated. The data is replaced with the calculated average value and output to the data bus. FIG.
In 2, the bidirectional half-pel interpolation unit 19 is provided in association with each texture memory. On the other hand, since each texture memory stores rendering data, a difference image, and a reference image, regardless of the type of these data,
Half-pel interpolation will be performed.

The sign bit determination unit 20 is a component provided exclusively for the first object front side texture memory 31 for storing the difference image in order to interpolate the difference point (1), and is used in the newly decoded difference image. The code bits of each pixel value are detected, and a code table 21 indicating all the pixels in the difference image and the code bits of those pixel values is generated in the memory. For pixel values, take the absolute value of the pixel value,
This absolute value is output to the data bus as the pixel value.

When the pixel value read from the first object front side texture memory 31 and subjected to half-pel interpolation by the bidirectional half-pel interpolation unit 19a is that of a difference image, the switch 37 converts the pixel value into a sign bit. Judgment unit 2
0, and if it is the first object front side rendering data, the output is switched so that the pixel value is directly output to the data bus.

Next, the internal structure of the synthesizing section 9 is shown in FIGS.
This will be described with reference to FIG. FIG. 17 to FIG.
FIG. 3 is a diagram showing an internal configuration of the device. Of these figures, FIG.
Corresponds to a state in which rendering data is stored in the first object front side texture memory 31 to the third object back side texture memory 35 shown in FIG. 15A, and FIG. 18 shows the state shown in FIG. This corresponds to the state after the P picture decoding process. FIG. 19 corresponds to the state after the decoding processing of the B picture shown in FIG. Or later,
How the internal configuration of the synthesizing unit 9 performs image synthesizing in the storage state of FIGS. 17 to 19 will be described.

The transmittance α 1 calculator 22a calculates the transparency set for the first object in which the front side rendering data is stored in the first object front side texture memory 31 and the first object
Selector 2 calculates the transmittance α1 of the first object based on the illuminance set for the vertex located on the front side of the object,
4a. The transmittance setting unit 23a is provided for the purpose of interpolating the difference point (2), and transmits 100% transmittance to the selector 24a for the pixel value of the difference image stored in the first object front side texture memory 31. Output (Fig. 18
reference).

When the rendering data is stored in the first object front side texture memory 31, the selector 24a outputs the transmittance calculated by the transmittance α1 calculator 22a to the multiplier 25a, and outputs the first object front side texture. Memory 3
When the difference image is stored in 1, the transmittance calculated by the transmittance α1 setting unit 23a is output to the multiplier 25a. The multiplier 25a multiplies the pixel value read from the first object front side texture memory 31 by the transmittance selectively output by the selector 24a to calculate (pixel value × α1%), and the multiplication is performed. The result is output to the adder 28a.

The transmittance α 2 setting unit 26 a determines the transparency set for the first object whose back side rendering data is stored in the first object back side texture memory 32,
The transmittance α2 of the first object is calculated based on the illuminance set for the vertex on the back side of the object and output to the selector 24a. The multiplier 27a calculates the pixel value read from the first object back side texture memory 32 and the transmittance α2 setting unit 2
6a is multiplied by the transmittance α2% output by the adder 28a.
Output to

Here, the reason why the transmittance setting section is not provided for the first object back side texture memory 32 storing the back side rendering data is that the difference image is a flat body, and only the front side data is provided. Because it does not exist. The transmittance α3 calculation unit 22b and the transmittance α4 calculation unit 22c are used to store the second object front side texture memory 33, the third
The transparency set for the second and third objects whose front side rendering data is stored in the object front side texture memory 35, the illuminance set for the front side vertices of the second and third objects, and The selectors 2 calculate the transmittances α3, α4 of the second and third objects based on the transmittances α1, α2 calculated for the first object located on the nearer side.
4b and c.

The transmittance α3 setting unit 23b and the transmittance α4 setting unit 23c are provided for the purpose of interpolating the difference (2), and the second object front side texture memory 33, the third object front side texture The transmittance of 50% of the pixel value of the reference image stored in the memory 35 is output to the selectors 24b and 24c (see FIGS. 19 and 20). The selectors 24b and 24c
When rendering data is stored in the second object front side texture memory 33 and the third object front side texture memory 35, the transmittance calculated by the transmittance α3 calculator 22b and the transmittance α5 calculator 22c is multiplied by a multiplier 25b. Multiplier 25c
To the second object front side texture memory 33, the third object
When the reference image is stored in the object front side texture memory 35, the transmittance set by the transmittance α3 setting unit 23b and the transmittance α5 setting unit 23c is output to the multipliers 25b and 25c.

The multipliers 25b and 25c are used to store the second object front side texture memory 33 and the third object front side texture memory 3
5 is multiplied by the transmittances selectively output by the selectors 24b and 24c and output to the adders 28b and 28c. The transmittance α4 calculation unit 26b and the transmittance α6 calculation unit 26c are set for the second and third objects in which the back side rendering data is stored in the second object back side texture memory 34 and the third object back side texture memory 36. Of the second and third objects based on the calculated transparency, the illuminance set for the rear-side vertices of the second and third objects, and the transmittance calculated for the object located on the near side from the viewpoint when viewed from the viewpoint. The transmittances α5 and α6 are calculated and output to the selectors 24b and 24c.

The multipliers 27b and 27c calculate the pixel value read from the first object front side texture memory 31 and the transmittance α
4 multiplies the values output by the calculator 26b and the transmittance α6 calculator 26c and outputs the result to the adders 28b and 28c. Adder 28
a is multiplied by the multiplier 25a when the front side and back side rendering data of the first object are stored in the first object front side texture memory 31 and the first object back side texture memory 32, respectively. The pixel value of the rendering data on the front side of the first object is added to the pixel value of the rendering data on the back side of the first object multiplied by the multiplier 27a and output to the signed adder 29. By such an addition process, a synthesized image is obtained in which the projected image on the back side is transparent from the projected image on the front side.
When the difference image is stored in the first object front side texture memory 31 and the first object back side texture memory 32 is not stored, the pixel value of the difference image multiplied by the multiplier 25a is added to the signed adder. 29.

When the front side and back side rendering data of the second object are stored in the second object front side texture memory 33 and the second object back side texture memory, respectively, the adder 28b is operated by the multiplier 25b. The pixel value of the rendering data on the front side of the multiplied second object and the rendering data on the back side of the second object multiplied by the multiplier 27b are added to add a signed adder 2
9 is output. When the reference image is stored in the second object front side texture memory 33 and the second object back side texture memory is not stored, the pixel value of the reference image multiplied by the multiplier 25b is added to the signed adder 29. Output to

If the front side and back side rendering data of the third object are stored in the third object front side texture memory 35 and the third object back side texture memory 35, respectively, the adder 28c multiplies by the multiplier 25c. The pixel value of the rendering data on the front side of the third object subjected to the addition and the rendering data on the back side of the third object multiplied by the multiplier 27c are added to add a signed adder 2
9 is output. If the reference image is stored in the second object front side texture memory 33 and the second object back side texture memory is not stored, the pixel value of the reference image multiplied by the multiplier 25c is added to the signed adder 29. Output to

The signed adder 29 further adds the result of addition by the adder 28a, the result of addition by the adder 28b, and the result of addition by the adder 28c, and outputs the result to the switch 41. First object front side texture memories 31 to 3
When the front side and back side rendering data are stored in the object back side texture memory 35, respectively, the addition result by the adder 28a is the pixel value of the front side rendering data of the first object and the pixel value of the back side rendering data. The result of addition by the adders 28b and 28c is the result of adding the pixel values of the front-side rendering data and the pixel values of the rear-side rendering data of the second object and the third object. By the addition of the adder 29, the rendering data for the three objects is synthesized. When the addition by the signed adder 29 is repeated for all the pixel values constituting the rendering data, the rendering data are combined as shown in FIG.

On the other hand, the first object front side texture memory 3
If the difference image and the reference image are stored in the first to third object back side texture memories 35, respectively, the adder 28a
Is the pixel value of the difference image, and the adder 28
b, since the addition result by the adder 28c is the pixel value of the reference image, the pixel value of the difference image and the reference image is added by the addition of the signed adder 29. At this time, for the difference image, the sign bit determination unit 20 generates a sign table according to the sign bit and replaces the pixel value with the absolute value, so that the sign bit of the difference image is reproduced. , The signed adder 29 refers to the code table and assigns a positive or negative sign to each pixel value of the difference image. As a result, when necessary, subtraction between the reference image and the difference image can be performed.

When the addition by the signed adder 29 is repeated for all the pixel values constituting the rendering data, the rendering data are combined with each other as shown in FIG. The saturation calculation unit 30
It is provided for the purpose of interpolating (1), and performs a positive value process and a saturation operation process on the addition result by the signed adder 29.

The positive value conversion process is a process of rounding the result of calculation by the signed adder 29 to a zero value or a positive value when the result is a negative number. Generally, the difference image is represented as signed data indicating a relative value with respect to the preceding and following data. Therefore, the product-sum value of each component of the compressed data and the predetermined coefficient may appear as a negative number. On the other hand, playback hardware such as a display and a speaker can process only unsigned data, so that when using a difference image, it is necessary to appropriately perform a positive value conversion process.

The saturation calculation process is a process of rounding a product-sum value to a predetermined value when the product-sum value exceeds a predetermined range (saturation). As described above, according to the present embodiment, 3
Display processing of three-dimensional graphics data and decoding of moving image data are displayed by adding a component for decoding moving image data to hardware provided for processing of displaying three-dimensional graphics data. Processing device 102
To store the image to be synthesized.
The object front side texture memory 31, the first object back side texture memory 32, and an adder for adding these pixel values can be shared. By this commonality, 3
It is possible to manufacture a multimedia playback device having both the display processing capability of three-dimensional graphics data and the decoding processing capability of moving image data at a lower price.

(Second Embodiment) In a second embodiment, the display processing device 102 performs a decoding process by a motion prediction process. For this motion prediction process, an accumulator for integration calculation is provided in the synthesizing unit 9. Have. In the second embodiment, when the display processing device 102 receives an instruction for decoding moving image data, the display processing device 102 writes one frame of frame data out of the plurality of frame data into the first object front side texture memory 31, and stores the frame data. Is written to the first object back side texture memory 32. When the two frame data are written, the pixel values of the frame data stored in the first object front side texture memory 31 are given a negative sign, and the pixel values of the frame data stored in the second memory are Is added to the signed adder 29 to obtain the difference information between the frame data and write it to the work memory 4.

When the writing to the work memory 4 is performed, the frame data to be displayed next to the frame data written to the first object back side texture memory 32 is written to the first object front side texture memory 31. 2
When the frame data is written, the pixel value of the frame data stored in the first object front side texture memory 31 is assigned a negative sign, and the pixel value of the frame data stored in the second memory is changed. The result is added to the signed adder 29 to obtain difference information between frame data. After obtaining the difference image in this way, the accumulator for integration calculation
The difference image stored in the work memory 4 is read out, the difference image newly obtained and the difference image previously written in the work memory 4 are integrated, and the integration result is written back to the work memory 4. A motion prediction process is performed based on the integrated value of the difference images obtained in this way, and a moving image is decoded.

As described above, according to the present embodiment, by adding the accumulator for integration calculation in the synthesizing unit 9,
Motion prediction processing can be realized with a simple configuration.
In the above-described embodiment, processes such as address calculation and determination for distinguishing between 3D graphic processing and motion compensation processing are duplicated for graphic processing and moving image data processing, thus providing redundancy. However, since these processes are all performed by the CPU executing the program, they do not lead to an increase in hardware resources, and do not cause the above effects to be impaired.

In the motion compensation processing of the P picture, both the difference image and the reference image are mixed as texture data. However, the difference image may be processed as frame data, and the reference image may be processed as texture data. Also, the code information of the difference image data is not provided by the method of providing the dedicated table shown here, but is provided, for example, by a difference bit that is present in the pixel value of the 3D graphic data and does not exist in the pixel value of the moving image data. Is also good. For example, 3D
If the pixel value of the graphic data is 24 bits and the pixel value of the moving image data is 16 bits, code information is set in a part of the upper 8-bit area that is the difference, and reference is made here. You may.

[0087]

According to the present invention, a plurality of rendering data indicating a plurality of projection images to be synthesized and displayed, and transparency information indicating how much each of the overlapping rendering data is transmitted at the time of synthesizing are obtained. Become 3
A display processing device that performs display processing on three-dimensional graphics data and performs display processing on moving image data including a plurality of frame data, and performs display processing on three-dimensional graphics data and display on moving image data. Receiving means for receiving an instruction for any one of the processes from the operator; receiving a display process instruction for the first memory, the second memory, and the three-dimensional graphics data; A first writing means for writing the data to be written into the first memory and writing the data to be the foreground to the second memory;
Blending means for blending a pixel value of the background-side rendering data written to the first memory and a pixel value of the foreground-side rendering data written to the second memory at a mixing ratio based on the transparency information; When an instruction for data display processing is received, difference frame data including only difference information indicating a difference from previous and subsequent frame data among a plurality of frame data, and reference frame data including only difference value data and a pixel value component to be added are included. And a second writing unit for writing the difference frame data and the reference frame data to the first memory and the second memory, respectively. The pixel value of the reference frame data written in the second memory and the pixel value of the difference frame data are Because characterized by blending at a constant mixing ratio of, when synthesizing a plurality of images,
Images to be synthesized can be stored in the common memory,
Further, the pixel values of the image to be combined can be added by a common adder. In particular, since the memory occupying a large area on the board and causing a rise in manufacturing cost can be shared, the effects of reducing the size and cost can be achieved.

In the above arrangement, the difference information is
Consists of signed data indicating the difference between pixel values for each pixel,
When receiving an instruction for the moving image data display processing, the second writing means writes the absolute values of all the pixel values constituting the difference information in the difference frame data to the first memory and generates an absolute value table in the memory. An absolute value table generation unit, and a code table generation unit configured to write code information indicating the sign of the pixel value of each pixel in another area in the memory to generate a code table in the memory. With such a configuration, even with a 3D graphic device that does not handle signed numeric values, it is possible to achieve an effect that signed difference image data for motion compensation processing can be processed.

In the above arrangement, the blending means has a signed adder for adding the pixel value of the background-side rendering data to the pixel value of the foreground-side rendering data. When a processing instruction is received, a reading unit that reads out all the absolute values from the absolute value table and attaches the sign of the pixel value indicated in the sign table to the absolute value of each pixel value read from the absolute value table A signed adder, wherein the signed adder is configured to perform addition of the signed pixel value and the pixel value of the reference frame data when the sign is assigned by the sign assigning unit. Often, in this case, even a 3D graphic device that does not handle signed numeric values can process signed difference image data for motion compensation processing.

[Brief description of the drawings]

FIG. 1 is a diagram showing an AV system having a multimedia playback device as a core.

FIG. 2 is a diagram showing a data structure of three-dimensional graphics data.

FIG. 3 is a diagram showing what shape each relative coordinate sequence specified in each object information represents.

FIG. 4 is a diagram showing what kind of pattern each piece of pattern information defined in each piece of object information represents.

FIG. 5 is a diagram showing how each coordinate in the three-dimensional graphics data has a positional relationship.

6 is a diagram showing a projection image of the three-dimensional graphics data shown in FIG.

FIG. 7 is a diagram showing a projected image to be displayed on a screen m5 when each coordinate in the three-dimensional graphics data has the positional relationship shown in FIG.

FIG. 8 is an explanatory diagram for comparing the size of slice data for each of an I picture, a P picture, and a B picture.

FIG. 9 is a diagram illustrating a state in which a plurality of rendering data are combined to generate a projection image.

FIG. 10 is a diagram showing a state in which a difference image and a reference image are combined to obtain a display image of the next frame.

FIG. 11 is a diagram showing the inside of a multimedia reproducing apparatus.

FIG. 12 is a diagram showing an internal configuration of the display processing device 102.

FIG. 13 is a diagram showing an internal configuration of a reading address generation unit 14;

FIG. 14 is a diagram showing how a base address generation unit 15 generates a base address of a read destination and a write destination when transmitting rendering data, a reference image, and a difference image.

15 (a) to (c) Base address generator 15
Is a diagram showing how rendering data, a reference image, and a difference image are stored in a first object front side texture memory 31 to a third object back side texture memory 35 by generating a base address.

FIG. 16 is a diagram showing components of the texture memory array 3 interposed between the first object front side texture memory 31 and the data bus.

FIG. 17 is a diagram showing an internal configuration of the synthesizing unit 9;

18 is a diagram illustrating a state of the combining unit 9 after the decoding processing of the P picture illustrated in FIG. 15B.

19 is a diagram illustrating a state of the combining unit 9 after the decoding processing of the B picture illustrated in FIG.

20 (a) to 20 (f) are diagrams showing what kind of rendering data is obtained by the icicle in the posture of FIG.

21 (a) and (b) are diagrams showing what rendering data is obtained by the pyramid in the posture of FIG.

FIGS. 22A to 22C show what first object front side rendering data, first object back side rendering data, and second object front side rendering data are stored in first object front side texture memory 31 and second object front side texture memory 33; It is a figure showing whether object front side rendering data is stored.

FIG. 23 is a diagram showing how a plurality of rendering data are combined to generate a projection image.

FIG. 24 is a diagram showing a projection image generated by the combination shown in FIG. 23;

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Input switching part 2 Frame memory 3 Texture memory array 4 Work memory 5 Geometry part 6 Slope calculation part 7 Shading part 8 Texture mapping part 8 Texture mapping part 9 Synthesis part 10 Decoding control part 11 Variable length decoding part 12 Dequantization Unit 13 Inverse DCT unit 14 Reading address generation unit 15 Base address generation unit 16 Rendering data offset address generation unit 17 Difference image offset address generation unit 18 Reference image offset address generation unit 19 Bidirectional half-pel interpolation unit 20 Code bit determination unit 21 Code Table 30 Saturation Operation Unit 200 Graphic Data Dependent Unit 200 3D Graphics Data Dependent Unit 201 Moving Image Data Dependent Unit

Continued on the front page F term (reference) 5B050 AA08 AA10 CA04 DA02 DA04 EA24 EA27 EA30 FA02 FA06 5B057 AA01 CA01 CA02 CA08 CA12 CA13 CA16 CB01 CB02 CB08 CB12 CB13 CB16 CC01 CD06 CE08 CH11 5B080 CA01 CA09 DA07 A08 BA08 A08 BB15 CA55 DA22 DA26 DA53 DA86 DA87 DA89 MM04 MM07

Claims (6)

[Claims] 1. A three-dimensional graphic comprising a plurality of rendering data indicating a plurality of projection images to be combined and displayed, and transparency information indicating how much each of the overlapping rendering data is transmitted at the time of the combination. A display processing device that performs a display process for moving image data and a display process for moving image data composed of a plurality of frame data, and a display process for three-dimensional graphics data and a display process for moving image data. Receiving means for receiving an instruction for one of them from the operator, a first memory, a second memory, and receiving a display processing instruction for the three-dimensional graphics data, the rendering data becomes a background side of the plurality of rendering data. To the first memory and the foreground side to the second memory First writing means for writing; and a pixel value of the background-side rendering data written to the first memory and a pixel value of the foreground-side rendering data written to the second memory at a mixing ratio based on the transparency information. Blend means for blending, and upon receiving an instruction of the moving image data display processing, difference frame data consisting only of difference information indicating a difference from previous and subsequent frame data among a plurality of frame data, and a pixel value to be added to the difference frame data And second writing means for writing reference frame data consisting only of components into the first memory and the second memory, respectively, wherein the blending means stores the difference frame data and the reference frame data in the first memory and the second memory. When written, the first memory, the second
A pixel value of the reference frame data written in the memory;
A display processing device, wherein pixel values of difference frame data are blended at a predetermined mixture ratio. 2. The difference information comprises signed data indicating a difference between pixel values for each pixel. When receiving an instruction for a moving image data display process, the second writing unit converts the difference information in the difference frame data. An absolute value table generating unit for writing the absolute values of all the pixel values to be configured in the first memory to generate an absolute value table in the memory; and sign information indicating whether the pixel value of each pixel is positive or negative in the memory. 2. The display processing device according to claim 1, further comprising: a code table generation unit that writes a code in an area and generates a code table in the memory. 3. The blending unit includes a signed adder that adds a pixel value of the background-side rendering data and a pixel value of the foreground-side rendering data. A reading unit that reads all absolute values from the absolute value table when an instruction is received, and a sign assignment that adds the sign of the pixel value indicated in the sign table to the absolute value of each pixel value read from the absolute value table And a sign adder for adding the sign-added pixel value and the pixel value of the reference frame data when the sign is assigned by the sign assigning unit. The display processing device according to the above. 4. The blending means replaces the result of addition by the signed adder exceeding a predetermined upper limit with an upper limit, and when the result of addition by the adder is a negative number, converts the addition result. The display processing device according to claim 3, further comprising a saturation operation unit that replaces the lower limit value. 5. The blending means includes: a multiplicand generating unit that generates a predetermined multiplicand; and a multiplier that multiplies the generated multiplicand by a pixel value to be added by a signed adder. Upon receiving an instruction to perform display processing on the three-dimensional graphics data, the transmissivity is calculated for each rendering data based on the transparency information included in the reference image, and the transmissivity is output to the multiplier as the multiplicand. Unit and a display processing instruction for the moving image data,
The display processing device according to claim 4, further comprising: a multiplicand calculating unit that outputs a predetermined multiplicand. 6. A display process for three-dimensional graphics data comprising a plurality of rendering data to be combined and displayed, and transparency information indicating how much each of the overlapping rendering data is transmitted at the time of the combination. And a coding processing for moving image data composed of a plurality of frame data, the display processing device comprising: a display processing device for three-dimensional graphics data; and a coding process for moving image data. Receiving means for receiving one of the instructions from the operator; a first memory, a second memory; and receiving a display processing instruction for the three-dimensional graphics data. First writing to write to memory and writing foreground side to second memory Weighting means for weighting a pixel value in the background-side rendering data and a pixel value in the foreground-side rendering data from the first memory with a value based on the transparency information; and adding the weighted pixel values to each other. Means, when receiving an instruction to decode moving image data, writes one frame of frame data out of the plurality of frame data into a first memory, and writes frame data to be displayed next to the frame data into a second memory. A second writing unit that, when the two frame data are written, assigns a negative sign to the pixel value of the frame data stored in the first memory and assigns the pixel value of the frame data stored in the second memory Adding means for obtaining the difference information between the frame data by adding the value and the adding means, and an absolute value for obtaining the absolute value of the difference information Means out, and integrating means for integrating the calculated absolute value, based on the integrated value of the absolute value, the display processing apparatus comprising: a generating means for generating a motion prediction data.