JP3556777B2 - Image processing device - Google Patents

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention, for example, displays a three-dimensional object in real time according to a user's input and a calculation result in limited hardware resources, although high accuracy is not required unlike a video game machine or a personal computer. The present invention relates to an image processing apparatus suitable as an apparatus required to perform the processing.
[0002]
[Prior art]
In a home TV game machine, a personal computer, a graphic computer, or the like, an image processing apparatus for generating image data to be output to a TV receiver, a monitor receiver, or the like for display is provided by a general-purpose memory, A high-speed processing is enabled by providing a dedicated drawing device between the CPU and the frame buffer as a flow of the drawing image data by combining a CPU and other arithmetic LSIs.
[0003]
That is, in the above image processing apparatus, when generating an image, the CPU does not directly access a frame buffer as a display memory corresponding to a display screen, but instead executes geometry conversion such as coordinate conversion, clipping, and light source calculation. Processing is performed to define a three-dimensional model as a combination of basic unit figures (polygons) such as a triangle and a quadrangle, and create a drawing command for drawing a three-dimensional image. Then, the CPU sends the drawing command to the drawing device via the external bus.
[0004]
The drawing command includes information such as the shape, position, orientation, color, and pattern of the polygon to be drawn. The shape, position, and orientation of a polygon are determined by the coordinates of its vertices.
[0005]
In the above-described image generating apparatus, for example, when displaying a three-dimensional object, the object is decomposed into a plurality of polygons, a drawing instruction corresponding to each polygon is generated by the CPU, and the generated drawing instruction is transmitted through a bus. Transfer to drawing device. The drawing device executes the drawing command, writes display data into the frame buffer, and displays a target three-dimensional object.
[0006]
At this time, in order to represent the object more practically, a method called texture mapping or mip mapping is used in which a predetermined image pattern is prepared and the inside of the polygon is modified by the image pattern.
[0007]
Further, a method of changing a display color by converting color data of an image through a color look up table (CLUT) in which color conversion data is recorded is widely known.
[0008]
[Problems to be solved by the invention]
By the way, the performance improvement such as the increase of the operating frequency and the reduction of the circuit scale of the operation LSI constituting the home TV game machine and the personal computer without increasing the cost is measured. The available general-purpose memory capacity has not increased much. For this reason, in home TV game machines and personal computers, the memory capacity is a so-called bottleneck.
[0009]
In particular, when performing texture mapping using a high-quality image (pre-rendering pattern) created in advance on a high-performance workstation as a texture pattern, it is necessary to store all patterns constituting a series of moving image segments in a memory. However, as the resolution of the texture pattern increases, the memory capacity is reduced accordingly.
[0010]
Therefore, the texture pattern is compressed and held in the memory, and the compressed texture pattern data is read from the memory each time it is used, and a dedicated image decompression device (a device for decompressing) is used. A method of expanding and decoding the data is used.
[0011]
In this case, in general, the decompression decoding time is much longer than the transfer time of the display image data to the frame buffer of the drawing device, so that the image decompression device cannot operate in synchronization with the drawing device. Therefore, the decompressed image is temporarily stored in a so-called main memory or a buffer memory such as a FIFO placed between the two devices so that the image decompression device and the drawing device can operate asynchronously. Since the data stored in the memory is data obtained by decompressing the compressed data, the memory capacity is greatly reduced.
[0012]
Also, with current circuit technology, signals can be transmitted at very high speed inside the LSI, but it is difficult to increase the signal transfer speed between the LSIs. In particular, in household equipment, the frequency of the transfer clock on the board cannot be increased. For this reason, even if processing is performed at high speed inside the LSI, the transfer speed on the bus for transmitting the result to another LSI often becomes a bottleneck. As described above, the decompressed image data must be temporarily stored in the main memory and transferred to the drawing device again. In this case, since all data is exchanged between the LSIs via the bus, In many cases, the processing speed at the stage becomes a bottleneck on the processing speed of the entire system.
[0013]
SUMMARY OF THE INVENTION In view of the above, it is an object of the present invention to provide an image processing apparatus which can increase the apparent transfer speed between LSIs via a bus and can improve the use efficiency of a memory.
[0014]
[Means for Solving the Problems]
In order to solve the above-described problems, an image processing device according to the present invention includes:
A memory, an image expansion decoding unit, and a drawing processing unit, wherein the compressed image data is decompressed and expanded by the image expansion decoding unit, and the expanded image data is transferred to the memory via a bus; An image processing apparatus, wherein the decompressed image data stored in a memory is transferred to the drawing processing unit via the bus, and display image data is generated by the drawing processing unit.
An instantaneous compression unit that instantaneously compresses the expanded image data is provided between the output end of the expanded image data of the image expansion decoding unit and the bus,
A decompression unit for decompressing the instantaneously compressed image data is provided between the input unit of the drawing processing unit and the bus.
[0015]
In the image processing device having the above configuration, the image data expanded by the image expansion decoding unit is instantaneously compressed by the instantaneous compression unit, and then transferred to the memory via the bus. At the time of transfer from the memory to the drawing processing unit, the decompressed image data that has been instantaneously compressed is decompressed in a decompression unit provided before the input unit of the drawing processing unit, and the decompressed image data obtained by the decompression is obtained. The image data is supplied to the drawing processing unit, and the drawing processing is executed.
[0016]
Since the decompressed image data transferred through the bus is data that is instantaneously compressed, the apparent transfer speed when viewed from an LSI such as a memory, an image decompression decoding unit, or a drawing processing unit is, for example, twice or more. Can be faster. Further, since the image data stored in the memory is compressed, the use efficiency of the memory increases.
[0017]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, an embodiment of an image processing apparatus according to the present invention will be described with reference to the drawings for a case of a video game machine.
[0018]
FIG. 2 shows a configuration example of a video game machine according to an embodiment of the present invention, which is an example of a game machine having a three-dimensional graphics function and a moving image reproduction function.
[0019]
FIG. 3 shows the appearance of the game machine of this example. The game machine of this example includes a game machine main body 1 and a control pad 2 constituting a user operation input unit. The control pad 2 is connected to the game machine main body 1 by coupling a connector plug 4 attached to a tip of a cable 3 connected to the control pad 2 to a connector jack 5A of the game machine main body 1. . In this example, two connector jacks 5A and 5B are provided in the game machine main body 1 so that two control pads 2 can be connected to the game machine main body 1 for a so-called battle game or the like. ing.
[0020]
The game machine of this example can enjoy the game by loading, for example, a CD-ROM disk 6 into the game machine body 1 as an auxiliary storage means in which a game program and image data are written.
[0021]
Next, the configuration of the game machine of this example will be described with reference to FIG. The game machine of this example has a configuration including two system buses including a main bus 10 and a sub bus 20. Data exchange between the main bus 10 and the sub bus 20 is controlled by the bus controller 30.
[0022]
The main bus 10 is connected to a main CPU 11, a main memory 12, an image decompression decoding unit 13, a preprocessing unit 14, a drawing processing unit 15, and a main DMA controller 16. A processing memory 17 is connected to the drawing processing unit 15, and the drawing processing unit 15 includes a frame buffer (frame memory) for display data and a D / A conversion circuit. Is output to the video output terminal 18. Although not shown, the video output terminal 18 is connected to, for example, a CRT display as a display device.
[0023]
The sub bus 20 includes a sub CPU 21, a sub memory 22, a boot ROM 23, a sub DMA controller 24, an audio processor 25, an input unit 26, an auxiliary storage unit 27, and a communication interface unit for expansion. 28 are connected. The auxiliary storage unit 27 includes a CD-ROM decoder 41 and a CD-ROM driver 42 in this example. The boot ROM 23 stores a program for starting up the game machine. The audio processing processor 25 is connected to an audio processing memory 25M. The audio processor 25 includes a D / A conversion circuit, and outputs an analog audio signal to an audio output terminal 29.
[0024]
The auxiliary storage unit 27 decodes application programs (for example, game programs) and data recorded on the CD-ROM disc 6 loaded in the CD-ROM driver 42. On the CD-ROM disk 6, for example, image data of moving images and still images compressed by the MPEG2 method using discrete cosine transform (DCT) and image data of a texture image for modifying polygons are also recorded. . The application program on the CD-ROM disk 6 includes a polygon drawing command.
[0025]
The input section 26 includes the control pad 2 as the operation input means described above, a video signal input terminal, and an audio signal input terminal.
[0026]
The main CPU 11 manages and controls each unit on the main bus 10 side. Further, the main CPU 11 performs a part of a process for drawing an object as a group of a large number of polygons. The main CPU 11 creates, on the main memory 12, a drawing instruction sequence for generating a drawing image for one screen, as will be described later. Data exchange between the main CPU 11 and the main bus 10 is performed in packet units by converting data into a packet format, thereby enabling burst transfer.
[0027]
The main memory 12 includes a memory area for compressed image data and a memory area for decompressed image data subjected to decompression processing for moving image and still image data. The main memory 12 has a memory area for graphics data such as a drawing instruction sequence (this is called a packet buffer). This packet buffer is used by the main CPU 11 for setting a drawing command sequence and transferring the drawing command sequence to the drawing processing unit.
[0028]
The image decompression decoding unit 13 decompresses the compressed moving image data reproduced from the CD-ROM disk 6 and transferred to the main memory and the compressed texture pattern data on the main memory 12. Since the image compression method described above is employed, a decoder structure corresponding to the image compression method is provided as described later.
[0029]
As will be described later, the output stage of the image decompression decoding unit 13 can perform instantaneous, that is, almost real-time, reversible compression / expansion, and has an instantaneous compression rate of, for example, about 1/4 to 1/2. A unit 50 is provided. In addition, the image decompression decoding unit 13 of this example uses a first output data format (hereinafter, this first output data) for requantizing and outputting the value of each pixel of the image data as an output format of the output image data. Format is referred to as a direct color format), and converts each pixel into index data indicating a color similar to the color of the pixel out of a predetermined limited number of reproduced colors, and outputs the converted data. An output data format (hereinafter, this second output data format is referred to as an index color format) can be selected in conformity with the processing in the drawing processing unit 15.
[0030]
The drawing processing unit 15 executes the drawing command transferred from the main memory 12, and writes the result to the frame memory. The image data read from the frame memory is output to the video output terminal 18 via the D / A converter and displayed on the screen of the image monitor.
[0031]
When the output format of the image data received from the main memory 12 is the index color format, the drawing processing unit 15 performs a process of converting each pixel data from the index data to the corresponding representative color data. For this purpose, the rendering processing unit 15 has a built-in CLUT (Color Look Up Table) which is a conversion table between index data and representative color data.
[0032]
The pre-processing unit 14 has a configuration of a processor including a CPU, and can share a part of the processing of the main CPU 11. For example, the pre-processing unit 14 may also perform a process of converting polygon data into two-dimensional coordinate data for display.
[0033]
In this example, an instantaneous decompression unit 60 that decompresses the instantaneous compression by the instantaneous compression unit 50 is provided between the preprocessing unit 14 and the main bus.
[0034]
Next, the basic processing of this game machine will be briefly described below.
[0035]
[Fetching of data from auxiliary storage unit 27]
When the power of the game machine in the example of FIG. 2 is turned on and the CD-ROM disk 6 is loaded in the game machine main body 1, a program for performing a so-called initialization process for executing the game in the boot ROM 23 is This is executed by the sub CPU 21. Then, the recording data of the CD-ROM disk 6 is taken in as follows.
[0036]
That is, in the auxiliary storage unit 27, the compressed image data, the drawing command, and the program executed by the main CPU 11 are read from the CD-ROM disk 6 via the CD-ROM driver 42 and the CD-ROM decoder 41, The data is temporarily loaded into the sub memory 22 by the sub DMA controller 24.
[0037]
The data fetched into the sub memory 22 is transferred to the main memory 12 by the sub DMA controller and the bus controller 30, and further by the main DMA controller 16. The sub CPU 21 is configured to directly access the frame of the drawing processing unit 15, and the sub CPU 21 can also change the display image content separately from the control of the drawing processing unit 15. It has been.
[0038]
[Decompression and transfer of compressed image data]
FIG. 1 is a block diagram for explaining in more detail mainly the flow of image data in the block diagram of FIG. In FIG. 1, dotted arrows indicate the flow of image data.
[0039]
As shown in FIG. 1, in this example, the image decompression decoding unit 13 includes a DMA controller 131, a FIFO memory 132, an MPEG decoder (hereinafter, referred to as MDEC) 133, a packer 134, a FIFO memory 135, A compression unit 50 is provided. The instantaneous compression section 50 includes a conversion table 52 for performing instantaneous compression, and a DMA controller 51.
[0040]
The DMA controllers 131 and 51 perform arbitration of the main bus 10 and divide the compressed image data and the decompressed image data (instantaneously compressed) into the main memory 12 and the image decompression device while the main bus 10 is idle. This is for DMA transfer with the decode 13. The FIFO memory 132 and the FIFO memory 135 are buffers with a minimum number of stages for preventing data from being lost even if a plurality of bus requests collide.
[0041]
The MDEC 133 decompresses and decodes data compressed according to the MPEG2 system. The decompression decoding process will be described by taking the case of texture pattern data as an example.
[0042]
Before explaining the decompression decoding, how the data of the texture pattern is compressed will be described.
[0043]
In this example, the texture pattern data is a two-dimensional image in which pixel data is composed of eight bits each of red (R), green (G), and blue (B) components. At this time, the pixel assigned to (R, G, B) = (0, 0, 0) is interpreted as a transparent color.
[0044]
First, the texture pattern is divided into a rectangular area of 16 pixels × 16 pixels as shown in FIG. This rectangular area is called a macro block. Thereafter, the texture pattern is processed on a macroblock basis. Macroblocks whose pixel values are all transparent are removed and packed in advance. At the same time, a table indicating the position information of the macroblock is recorded in the header that is the additional information.
[0045]
As shown in FIG. 4, the macroblock is converted from a three-primary-color signal expression into an expression composed of a luminance signal and a color difference signal. This is hereinafter referred to as CSC (Color Space Cover). s Ion). As shown in FIG. 4, the macro block composed of the luminance components is divided into four parts, and is composed of four luminance blocks Y0, Y1, Y2, and Y3 each having an area of 8 × 8 pixels. In addition, the macroblock composed of the color difference signal components is composed of two color difference signal blocks in which four adjacent pixels are grouped into a block of 8 pixels × 8 pixels. Thus, the macroblock is divided into a total of six blocks.
[0046]
For this macro block, DCT (Discr e te Cosine Transformation). DCT is a kind of similarity transformation generally called orthogonal transformation. Considering a matrix X of 8 × 8 pixels having each luminance value of a block as a component, using a DCT matrix P and its inverse matrix Pi,
Y = P ・ X ・ Pi
Refers to a transformation defined in the form The coefficients of the DCT matrix P are as shown in Table 1 shown in FIG.
[0047]
The DCT-transformed block is quantized at a different resolution for each component. A table that specifies the quantization width (step) for each component is called a quantization table (Q table). An example of the quantization table is shown in Table 2 of FIG.
[0048]
The actual quantization is performed by dividing the value of the Q table corresponding to each component by the product of the value QUANT that determines the overall quantization step.
[0049]
Here, if the value of the whole quantization step QUANT is increased, the image quality after decoding is degraded, but the number of 0 components in the block increases, so that the compression ratio can be improved.
[0050]
After the quantized blocks are numbered one-dimensionally in an order called zigzag order, variable-length coding by Huffman coding is performed. At the same time, in this example, a 16 × 16 × 1 bit α pattern (mask pattern) in which the bit corresponding to the transparent color becomes 1 for each macroblock.
[0051]
The result obtained in a macroblock unit is the final compressed image format, which is called a bit stream. That is, the continuation of a set of data and a mask pattern in units of macroblocks is data of a texture pattern of 64 × 64 pixels, which is recorded on the CD-ROM 6. Further, in the CD-ROM, a table indicating the above-described macroblock position information is also recorded in the header as the texture pattern data as described above.
[0052]
The procedure of image expansion decoding is performed by following the reverse of the above-described image compression procedure. Since the compression in this case is irreversible, the decoding is performed using a mask pattern that is interleaved and recorded in the pixel value compressed data so that the transparent color pixels of the decompressed macroblock are correctly decoded. The pixel of which the bit of the corresponding mask pattern is 1 is forcibly changed to a transparent color regardless of the pixel value of the decoding result.
[0053]
The MDEC 133 is configured by the following processing.
(1) The compressed image and additional information (geometry information and the like) are interleaved and recorded on the CD-ROM 6. These pieces of information are continuously read out by the CD-ROM driver 42 and the CD-ROM decoder 41, and once stored in the main memory 12 as described above, only the compressed image information is cut out. Then, it is transferred to the image decompression decoding unit 13. The additional information is processed in the CPU 11, whereby the position information of the object whose decompressed image is used as the texture is calculated.
(2) The MDEC 133 includes a variable length decoder, and decodes the Huffman-encoded block. The Huffman tree at this time is fixed, but the value of the corresponding code can be changed.
(3) The MDEC 133 also includes an inverse quantizer. The inverse quantizer performs inverse quantization on the decoded block, and converts the block order according to the above-described zigzag order. The inverse quantization is performed in different steps for each coefficient unit of the block.
(4) The inverse DCT converter included in the MDEC 133 performs an inverse orthogonal transform of 8 × 8 pixels.
(5) The MDEC 133 further includes a CSC processing unit, and converts a macroblock image represented by a luminance signal / color difference signal into an image represented by three primary color signals R, G, and B. An example of the conversion is shown in Table 3 in FIG.
[0054]
As described above, of the input data transferred from the auxiliary storage unit 27 to the main memory 12, the compressed image data is transferred from the main memory 12 to the image decompression decoding unit 13 by the DMA controller 131, and the MDEC 133 Is performed, the compressed image data is decompressed, and each pixel is decoded as direct color image data composed of the three primary color signals R, G, and B as described above.
[0055]
This image data is supplied to the packer 134. The packer 134 packs the decompressed and decoded image data into a format suitable for the drawing processing unit 15 on a pixel-by-pixel basis. That is, as described above, in this example, the output format of the image data to be sent to the drawing processing unit 15 can be either a direct color format or an index color format. 134 is performed.
[0056]
This packer 134 has a configuration as shown in FIG. That is, the packer 134 performs dither processing based on the dither matrix table 72, and vector quantization for converting pixel data into a representative color prepared in the CLUT 74 in order to convert the pixel data into the index color format. And a pack processing unit 75.
[0057]
When outputting in the direct color format and the number of bits of the input pixel is equal to the number of bits of the output pixel, the dither processing unit 71 and the vector quantizer 73 are bypassed, and the decoded data It is packed and output in units.
[0058]
Even in the case of outputting in the direct color format, when the bit number N of the output pixel is smaller than the bit number of the input pixel, the dither processing unit 71 performs an appropriate rounding process. In the case of this example, decoded data is obtained from the MDEC 133 in a 16-bit signed fixed-point format, and the rounding process is one of the following.
a) After the clipping process is performed so that the input pixel falls within the range of N bits, the lower N bits of the integer part of the input pixel value are output.
b) The upper N + 1 bits of the input pixel are rounded off and the upper N bits are output.
c) After the input pixel is multiplied by the ordered dither of the fixed dither matrix table 72, the upper N + 1 bits are rounded off and the upper N bits are output.
[0059]
Next, when outputting in the index color format, vector quantization using a pixel value (pixel color data) preset in the CLUT 74 as a representative point is performed in the vector quantizer 73, and each pixel value is output. Instead, the data of the index of the corresponding representative point is output.
[0060]
That is, FIG. 9 shows an example of the CLUT 74, which is an example in a case where there are 256 representative colors. That is, 256 color data 0 to color data 255 are selected in advance as possible colors of an image, registered as the CLUT 74, and an index is given as a color number of each color data.
[0061]
As a method of vector quantization, for example, pixel values after decompression decoding (for example, 8-bit color data for each color component of the three primary colors R, G, and B) and 16 to 256 types given as the CLUT 74 are used. (For each color component of the three primary colors R, G, and B, for example, consisting of 8 bits), and in the CLUT 74, the index of the color data having the closest color is used instead of the pixel value. Output.
[0062]
In this case, the proximity (distance) d between the two colors (R1, G1, B1) and (R2, G2, B2) is calculated by the following formula, for example.
[0063]
d = (R1-R2) * (R1-R2) + (G1-G2) * (G1-G2) + (B1-B2) * (B1-B2) (Q1)
In this formula (Q1), * indicates multiplication.
[0064]
As described above, in the index color method, the 24-bit pixel value after decompression decoding is compression-converted into 8-bit index data and output.
[0065]
As described above, the pixel data packed by the packer 134 is sent to the instantaneous compression unit 50 through the FIFO memory 135, and the image data is instantaneously compressed. The instantaneous compression here differs from high-efficiency compression such as MPEG2 in that the compression ratio is low at 1/4 to 1/2, but is reversibly compressed / decoded at high speed by a compression / decoding circuit with a small hardware scale. What can be used is used.
[0066]
In this example, run-length coding and Huffman coding are performed simultaneously for compression. A conversion table 52 which is a codebook as a dictionary for this compression is provided in the instantaneous compression unit 50. The conversion table 52, which is a codebook, is generated and held in advance.
[0067]
As shown in the figure, the instantaneous compression unit 50 functionally includes a DMA controller 51 and uses a conversion table 52 to simultaneously execute run-length encoding and Huffman encoding to instantaneously compress an image that has been MPEG expanded and decoded. The data is transferred to the main memory 12. The above is the operation of the image expansion decoding unit 13.
[0068]
When performing motion compensation, the instantaneous lossless compression in the instantaneous compression unit 50 is not performed. In this case, when reading the bit stream, the image decompression decoding unit 13 simultaneously reads and processes the image data of the previous frame expanded on the main memory 12.
[0069]
When a certain amount of unit data called macroblocks of the decompressed image data is stored in the main memory 12, the main CPU 11 transmits the decompressed data to the instantaneous decompression unit 60 and the preprocessing unit 14 through the main bus 10. Then, the image data is transferred to the frame buffer of the drawing processing unit 15 via the CPU. At this time, if the decompressed image data is transferred to the image memory area of the frame buffer, it will be displayed as it is on the image monitor device as a background moving image. If the image data is transferred to the texture area of the frame buffer, the image data of this texture area is used as a texture image for modifying the polygon.
[0070]
The instantaneous decompression unit 60 includes a DMA controller 61 as a functional block, and a conversion table 62 for performing an inverse conversion of the conversion table 52 of the instantaneous compression unit 50. Is decompressed using the conversion table 62, and is converted into MPEG-decompressed decoded image data, which is sent to the drawing processing unit 15 through the preprocessing unit 14.
[0071]
In the case of this example, when the output format is the direct color format, the pre-processing unit 14 renders image data consisting of R, G, and B primary color signals of a predetermined number of bits designated by each pixel data. The image data is supplied to the processing unit 15 and the drawing processing is executed.
[0072]
On the other hand, in the case of the index color format, the drawing processing unit 15 is supplied with the above-described index data. As described above, the drawing processing unit 15 includes a CLUT similar to the above-described CLUT 74, performs processing of converting image data in index color format into corresponding representative color data, and restores the image data. Then, a drawing process is executed using the restored image data.
[0073]
[Processing and transfer of drawing instruction sequence]
The three-dimensional image is displayed on the two-dimensional image display surface by drawing the polygons constituting the surface of the object in order from the polygon at the deep position in the depth direction according to the Z data which is the three-dimensional depth information. Can be. The main CPU 11 creates, on the main memory 12, a drawing command sequence for causing the drawing processing unit 15 to perform drawing in order from the polygon located at a deep position in the depth direction.
[0074]
The main CPU 11 calculates the movement of an object or a viewpoint based on a user's operation input from the control pad of the input unit 26, and creates a polygon drawing command sequence on the main memory 12.
[0075]
When the drawing command sequence is completed, the main DMAC 16 transfers the drawing command from the main memory 12 to the drawing processing unit 15 for each drawing command through the preprocessing unit 14.
[0076]
The drawing processing unit 15 sequentially executes the transmitted data and stores the result in a drawing area of the frame memory. At the time of this polygon drawing, the data is sent to the gradient calculation unit of the drawing processing unit 15 and the gradient calculation is performed. The gradient calculation is a calculation for obtaining the inclination of the plane of the mapping data when filling the inside of the polygon with the mapping data by polygon drawing. In the case of texture, polygons are filled with texture image data, and in the case of Gouraud shading, polygons are filled with luminance values.
[0077]
In addition, animation textures are possible. That is, in the case of the moving image texture, the compressed moving image data from the CD-ROM disk 6 is once read into the main memory 12 as described above. Then, the compressed image data is sent to the image expansion decoding unit 13. The image data is expanded by the image expansion decoding unit 13.
[0078]
Then, as described above, the expanded moving image data is sent to the texture area on the frame memory of the drawing processing unit 15. Since the texture area is provided in the frame buffer of the drawing processing unit 15, the texture pattern itself can be rewritten for each frame. As described above, when a moving image is sent to the texture area, the texture is dynamically rewritten and changed every frame. If texture mapping to polygons is performed using the moving image in the texture area, the texture of the moving image is realized.
[0079]
As described above, according to this embodiment, the decompressed image data is instantaneously compressed and transferred to the main memory via the main bus 10. Therefore, the use efficiency of the memory is improved by the amount of compression. Further, when the output format of the decompressed image data is the index color format, the pixel data is composed of index data, so that the data amount is reduced, and the use efficiency of the main memory is improved accordingly.
[0080]
The decompressed image data transferred from the image decompression decoding unit 13 to the main memory 12 via the main bus 10 and the decompressed image data transferred from the main memory 12 to the drawing processing unit 15 via the main bus 10 are respectively instantaneously compressed. Data, and the amount of data is small, so that the transfer speed of the bus is improved.
[0081]
Further, in this embodiment, even if the format of the input compressed image data is one type, the output format of the output data from the image decompression decoding unit 13 is selected from the direct color format and the index color format. Since it is not necessary to separately prepare input image data in order to obtain different output formats, the use efficiency of the main memory is improved in this respect as well.
[0082]
Even in the case of the direct color format, the number of bits suitable for the processing in the drawing processing unit 15 can be rounded by dither processing, so that output data of a desired number of bits can be easily obtained.
[0083]
Although the above example is a case where the image processing apparatus according to the present invention is applied to a game machine, it goes without saying that the image processing apparatus according to the present invention can be used for various purposes.
[0084]
【The invention's effect】
As described above, according to the present invention, the apparent transfer speed of decompressed image data transferred on the bus is improved, and the amount of decompressed image data stored in the memory can be reduced. Efficiency is improved.
[0085]
In addition, since only one type of compressed image data is required regardless of a plurality of output formats of decompressed image data, there is no need to store a plurality of compressed image data for a plurality of output formats in a memory. There is also an effect that the memory efficiency is improved.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating an example of a processing portion of image data in an embodiment of an image processing apparatus according to the present invention.
FIG. 2 is a block diagram illustrating a configuration example of a game machine as an embodiment of the image processing device according to the present invention.
FIG. 3 is a diagram showing an example of an appearance of the game machine in the example of FIG. 2;
FIG. 4 is a diagram for explaining a processing unit of compressed image data according to the embodiment of the present invention;
FIG. 5 is a diagram for explaining a compression method of compressed image data according to the embodiment of the present invention.
FIG. 6 is a diagram for explaining a compression method of compressed image data according to an embodiment of the present invention.
FIG. 7 is a diagram for explaining a compression method of compressed image data according to an embodiment of the present invention.
FIG. 8 is a block diagram showing a detailed example of some blocks in FIG. 1;
FIG. 9 is a diagram for explaining an index color format.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 10 ... Main bus, 11 ... Main CPU, 12 ... Main memory, 13 ... Image decompression decoding part, 14 ... Preprocessing part, 15 ... Drawing processing part, 16 ... Main DMA controller, 50 ... Instantaneous compression part, 51 ... Instantaneous compression Conversion table, 60: instantaneous decompression unit, 61: conversion table for compression / decompression, 71: dither processing unit, 72: dither matrix table, 73: vector quantizer, 74: CLUT, 75: pack processing unit, 132 And 135 ... FIFO memory, 133 ... MPEG decoder, 134 ... Packer

Claims (4)

A memory, an image decompression decoding unit, a drawing processing unit, and a data transfer control unit, wherein the compressed image data is decompressed and decompressed by the image decompression decoding unit, and the decompressed image data is transferred to the data transfer unit. Transferred to the memory through a bus under the control of the control unit, the decompressed image data stored in the memory is transferred to the drawing processing unit via the bus, and display image data is generated by the drawing processing unit. An image processing device,
A compression unit that compresses the decompressed image data almost in real time is provided between the output terminal of the decompressed image data of the image decompression decoding unit and the bus,
A decompression unit that decompresses the compressed image data is provided between the input unit of the drawing processing unit and the bus,
A first output data format for requantizing and outputting a value of each pixel of the expanded image data when transferring the expanded image data to the memory; and Can be selected from among a predetermined limited number of reproduction colors into index data indicating a color similar to the color of the pixel and outputting the index data. the image processing apparatus according to claim <br/> that is.   A memory, an image decompression decoding unit, a rendering processing unit, and a data transfer control unit, wherein the compressed image data is decompressed and decompressed by the image decompression decoding unit, and the decompressed image data is transferred to the data transfer unit. Transferred to the memory via a bus under the control of the control unit, the decompressed image data stored in the memory is transferred to the drawing processing unit via the bus, and display image data is generated by the drawing processing unit An image processing device,
  A compression unit that compresses the decompressed image data almost in real time between the output end of the decompressed image data of the image decompression decoding unit and the bus;
  A first output data format for requantizing and outputting a value of each pixel of the expanded image data when transferring the expanded image data to the memory; and Can be selected from among a predetermined limited number of reproduced colors into a second output data format that converts and outputs index data indicating a color similar to the color of the pixel. And
  The compression unit does not compress the decompressed image data when performing motion compensation on a display image.
An image processing apparatus characterized by the above-mentioned. The image processing apparatus according to claim 1 or 2,
When the first output data format is selected, quantization means for requantizing the value of each pixel of the decompressed image data using a dither method is provided on the output side of the image decompression decoding unit. An image processing apparatus characterized in that: The image processing apparatus according to claim 1 or 2,
When the second output data format is selected, quantization means for vector-quantizing the decompressed image data in accordance with the predetermined number of reproduction colors is provided on the output side of the image decompression decoding unit. An image processing apparatus, comprising:

Images