JP2937212B2 - Data processing device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for processing data from a CD-ROM on which data such as image data is recorded.

[0002]

2. Description of the Related Art A CD-ROM has a large recording capacity and is used as an external storage medium in a game machine or a personal computer using a microcomputer.
It is considered that the image data is recorded, the image data is read and supplied to a host computer, and a moving image is displayed on a CRT display.

In this case, in a conventional game machine, a CD-ROM decoder is connected to a single system bus connected to a CPU serving as a host computer of the game machine. As a method for reading image data from the CD-ROM, image data is not read continuously from the CD-ROM and a moving image is displayed on a display, but is recorded in a predetermined recording area of the CD-ROM. The image data of a moving image composed of a plurality of frames is read and transferred to a large-capacity RAM provided in the host computer, and an animation is created and displayed from the image data of a plurality of frames stored in the RAM. I have.

[0004]

However, according to this method, the number of frames per second of the moving image is small,
Cannot get smooth animation of movement.

Therefore, it is conceivable that a moving image is reproduced in real time while continuously reading out image data from a CD-ROM to realize an animation with a large number of frames and a smooth motion.

However, this cannot be realized by a configuration in which a CD-ROM decoder is connected to one system bus to which a CPU and a RAM as a buffer memory are connected as in a conventional game machine.

That is, since the data transmission rate of the CD-ROM is 150 Kbytes / sec, in order to display a moving image, the image data of the moving image is compressed and the CD is read.
-Recorded in a ROM, and at the time of display, the compressed image data is reproduced, and it is necessary to decode the original image data before supplying it to the display. Number) is insufficient, and it is not possible to display a sufficiently moving video. However, with the above-described conventional configuration, data cannot be processed at a processing speed capable of displaying a moving image having sufficient motion.

Further, in order to display a moving image using such image data, it is usually necessary to read image data from a CD-ROM and synchronize with a host computer. However, if such synchronization is taken, the configuration of the system becomes complicated. The present invention seeks to solve the above problems.

[0009]

In order to solve the above problems, according to the present invention, referring to the reference numerals of the embodiments described later, the processing apparatus main body 1 and the compressed image data are recorded.
The data is reproduced from the recording medium 5, decoded and compressed.
Made from the sub-processing unit 4 for obtaining image data, the processing device
The main body 1 is accessed by the CPU 11 and the CPU 11
And DMA transfer between each other is possible.
A number of system bus 18 and 19, the first buffer memory <br/> Li 13 connected to one bus 18 of the two system bus, to the other bus 19 of the two system bus
Connected to the data processor 14 and the second
A data processing unit having a buffer memory 15 ;
The sub-processing unit 4 is the other of the two system buses.
It is connected to the system bus 19, Deco by the sub-processing unit 4
The compressed image data obtained by loading
DMA transfer to the buffer memory 13 and the first buffer
The compressed image data from the memory 13
DMA transfer to the second buffer memory 15 of the processing unit
The compressed image is processed by the data processor 14.
Provided is a data processing device characterized by decompressing and decoding image data .

A data processor having a data processor 14 and a buffer memory 15 is connected to the system bus 19, and data from the CD-ROM decoder 43 is DMA-transferred to the first memory. And a data processor for DMA-transferring data from the second memory to the second memory and performing data processing by a data processor.

[0011]

The sub-processing section 4 is a first buffer of the processing apparatus main body 1 .
System bus 19 of those who § memory 13 is not connected
It is connected to the. Therefore , recording from the sub-processing unit 4
The compressed image data reproduced and decoded from the medium can be DMA-transferred to the memory 13. Then , the compressed image data is transferred from the memory 13 to the system bus 19.
Buffer memory of the data processing unit connected to the
15 is DMA-transferred, in the data processor, wherein
The compressed image data is decompressed and decoded. Thus, record
Decompression Deco compressed image data from the medium 5 in real time
Can be loaded . Therefore, the recording medium 5
By processing these moving image data, an animation of a sufficient number of frames can be displayed.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a data processing apparatus according to the present invention will be described below, taking as an example a case where animation data can be displayed in real time by reproducing moving image data recorded on a CD-ROM.

First, a method for compressing moving image data and a method for recording compressed moving image data and other data other than moving images on a recording medium will be described.

[Data Compression Method for Moving Picture Image Data] FIGS. 4 and 5 are block diagrams of an example of an encoding apparatus that executes the picture data compression method of this embodiment. In this example, the compressed image data is recorded on a CD-ROM. This CD-ROM is used as software for a game machine as described later, and image data is highly efficiently compressed so that a moving image can be reproduced.

In this example, one frame (one screen)
Is, as shown in FIG. 6A, horizontal × vertical = 256 pixels × 192
Each pixel is represented by five bits for each of the three primary colors. Actually, one bit of a dummy is added to the highest order for the sake of processing, and one pixel is set to one bit (dummy) +5 bits × 3 colors, that is, 16 bits. Then, the original image data is subjected to data compression processing in units of one frame as follows.

That is, the data of one frame of the original image is supplied to the character dividing means 22 through the input terminal 21, and as shown in FIG. 6B, the image of one frame is composed of horizontal × vertical = 8 pixels × 8 pixels. It is divided into small area blocks (this block is hereinafter referred to as a character).
Therefore, as shown in FIG. 6B, an image of one frame is divided into 32 × 24 = 768 characters. The image data C (0) to C (76) of each character
7) is temporarily stored in the register 23.

The image data C (0) to C (767) of each character from the register 23 is supplied to the first vector quantization means 24. Also in this example, in the vector quantization means 24, image data C (0) to C (767) of each character are processed in parallel. Of course, the image data C (0) to C (767) may be sequentially vector-quantized without performing the parallel processing. The same applies to each processing described later.

The vector quantization means 24 performs vector quantization for each character image data C (k) (k = 0 to 767) such that the number of colors represented as pixels in the character is within four. Done. As a method of this vector quantization, various proposals can be used,
In this example, when considering a three-dimensional color space in which the color components of the three primary colors of red, blue, and green are orthogonal to each other, a distance in the color space between each pixel is obtained, and pixels having a short distance from each other are determined. The pixel data is rounded so that the colors of the pixels in the character fall within four or less representative colors by grouping, that is, by performing processing of grouping pixels of similar colors into one representative color.

Then, for all the characters in one frame, vector quantization is performed so that the colors of the pixels in the character fall within four colors, and then the quantization error (the position of the representative color) in all the characters in the one frame is obtained. , The maximum value Emax of the representative color and each pixel in the color space is determined. At this time, a threshold value Eth allowed as the maximum value of the quantization error in one frame is set in advance. Then, the maximum value Emax of the quantization error is compared with a threshold value Eth. When the maximum value Emax of the quantization error is larger than the threshold value Eth, the quantization error is further reduced for the image data in each character.
Vector quantization is performed until just before the number of colors exceeds the number of colors in the character. This is to make the S / N of the image data in all the characters in one frame uniform. This is performed until immediately before the maximum value Emax of the quantization error exceeds the threshold value Eth. If you do this,
The S / N ratio for all frames is kept constant.

When quantization is performed in this manner, the number of pixel colors is reduced in a character having a flat color change. This is because, for a character having a flat color change, the quantization error does not increase so much even if the number of colors decreases. In this process, some characters have two colors in the character and only one color. Then, the color selected in each character is set as the representative color.

Thus, from the vector quantization means 24, image data compressed to four colors or less in each character is obtained. The image data in character units from the vector quantization means 24 is supplied to the palette dividing means 25.

The palette dividing means 25 classifies the characters into eight groups (each group is referred to as a palette) by grouping characters having similar colors according to the distribution of colors in the characters. For example, as shown in FIG. 6C, when a group of characters having similar colors according to the contents of the image occurs as A, B, C, D, E,..., The groups A, B, C, D,
A pallet is constructed for each E.

In the case of this example, the method of allocating eight palettes is as follows: (1) It is assumed that the representative color of each character (the average value of the colors in the character) is calculated, and that each character is composed of the representative color. (2) Vector quantization is performed, and all characters in one frame are quantized into eight colors. That is, since the number of characters is 768, the representative color of the character is a maximum of 768.
The colors are quantized into eight-color characters. (3) Characters having the same label (representative color) are combined into one palette. This is performed in three steps.

The group of characters constituting the pallet does not need to be in a continuous character area, but a group of characters that are discrete may constitute one pallet group.

Eight pallet data P (0) to P (7)
Are temporarily stored in a register 26, respectively, supplied to a second vector quantization means 27, and subjected to parallel processing.

The second vector quantization means 27 determines representative colors of 16 pixels for each pallet. At this time, if the number of colors of the pixels in one palette is larger than 16 colors, vector quantization is performed and the colors in the palette are rounded to 16 colors in the same manner as in the character. Then, the resulting 16 colors are set as the representative colors of the pixels.

Thus, 8 rounded to 16 colors each
Image data P (0) to P for each pallet in character units
(7) are supplied to the labeling means 28 and are processed in parallel. In each of the labeling means 28, the color conversion tables COL (0) to COL (0) to 16 colors or less than 16 color data selected as the representative colors of the pixels for each palette.
OL (7) is created and temporarily stored in the register 29 (see FIG. 7). The data of the color conversion tables COL (0) to COL (7) from the register 29 is supplied to the recording processing means 38 as recording data.

Each labeling means 28 refers to each of the color conversion tables COL (0) to COL (7), and calculates 16 characters for each character included in each palette.
The pixel data rounded to a color is converted into label image data LAB (0) to LAB (7) in which the color of the pixel is represented by a corresponding color number on the color conversion table of the palette (FIG. 8). reference). Then, the label image data LAB
(0) to LAB (7) are temporarily stored in the register 30.

In this case, as described above, there are characters composed of four or three colors (FIG. 8A), characters composed of two colors (FIG. 8B), and characters composed of only one color (FIG. 8C). If the character is 4 or 3 colors, the 4 or 3
If there is a table indicating the color number of a color, each pixel data can be represented by 2-bit data indicating which of the color number tables. Therefore, each pixel data of a character composed of four or three colors can be represented by two bits. Similarly, if the character is two colors,
Two color number tables for the character
It can be represented by bit pixel data. Further, if there is only one color, only the color data can be used as described later.

A character that can be represented by 2 bits is referred to as a 2-bit mode character, a character that can be represented by 1 bit is referred to as a 1-bit mode character, and a character having only one color is referred to as a single color character.

In consideration of the decoding processing, high-speed processing can be performed by treating the 2-bit mode character, the 1-bit mode character, and the single-color character collectively. However, in the 768 characters in one frame, generally, each mode character is dispersedly mixed as shown in FIG. 9A. In FIG. 9, is a 1-bit mode character, is a 2-bit mode character,
○ indicates a single-color character.

Then, the label image data LAB (0) to LAB (7) of each pallet from the register 30 are supplied to the sorting means 31, and as shown in FIG. 9B, a 2-bit mode character, a 1-bit mode character, One frame of character data is rearranged in the order of monochromatic characters.

In the sorting means 31, a table (hereinafter referred to as a screen table) s for rearranging the characters of one frame into the original order is provided.
cr is formed. This screen table scr is
As shown in FIG. 10, when an image of one frame is divided into small areas having the same size as a character, a character number CNo. And a pallet number PNo. The character number CNo. Is the character order in one frame after sorting the characters to be displayed at the position of the small area. The pallet number PNo. Indicates in which of the eight pallets the character displayed in the small area is included. That is, it indicates which color conversion table is used at the time of decoding. In this case, the character number CNo. And the pallet number PNo. Of one small area are composed of, for example, 2-byte data.

In the case of this example, the character number CN
The numbers 0 to 15 in o. are assigned only to monochrome characters. That is, in the table scr, when the character displayed at the position of a certain small area is a single color character, the palette number PNo. Is assigned to the small area in the same manner as the 2-bit mode or 1-bit mode character. , Character number C
Instead of No., 0 to 0 of the color conversion table of the palette
The color number of the color of the character among the 15 color numbers is assigned. Thereby, the color (single color) of the small area is determined. Therefore, for a single-color character, by registering and recording the color data of the character in the screen table scr as described above, it is not recorded as compressed image data for each character described later.

For the above-described single-color characters, the character numbers for the characters in the 2-bit mode and the 1-bit mode start from the 16th character. Originally, 10 bits are assigned to the character number,
There is ample room for such number shifts.

The data of the screen table scr is supplied to the recording processing means 38 as recording data.

Then, as described above, the sorting means 31
In the character-based image data sorted and rearranged in (1), N (N is an integer of 768 or less) character data C2 (0) to C2 (N-1) of each 2-bit mode character
Is supplied to the labeling means 33 via the register 32. In the labeling means 33, as shown in FIG. 11A, for each of the data C2 (0) to C2 (N-1) of the character in the 2-bit mode, a color number table of four or three colors of the character, Compressed image data dat2 (0) to dat2 (N-1) are formed from data of a 2-bit index number indicating each color number position on the color number table. Then, the compressed image data dat2 (0)
~ Dat2 (N-1) is temporarily stored in the register 34.

Similarly, M (M is 7)
The data C1 (0) to C1 (M-1) of each 1-bit mode character (an integer of 68 or less) are supplied to the labeling means 36 via the register 35. In the labeling means 36, character data C1 of each 1-bit mode
For (0) to C1 (M-1), as shown in FIG. 11B, a color number table of two colors of the character and data of a 1-bit index number indicating each color number position on the color number table are obtained. Compressed image data dat1 (0) to da
t1 (M-1) is formed. Then, the compressed image data dat1 (0) to dat1 (M-1) are temporarily stored in the register 37.

Then, all the compressed image data in the 2-bit mode from the register 34 and all the compressed image data in the 1-bit mode from the register 37 are supplied to the recording processing means 38 as recording data.

Data such as resident characters and game condition changes are supplied to the recording processing means 38 through a terminal 39.

[Generation of Recording Data] The recording processing means 38 creates data to be recorded on a CD-ROM. In this example, one frame is processed as one block in this recording data, but the data recording mode on the CD-ROM is a CD-ROM.
Of course, it follows the data format of the ROM.

For example, the sectors when the recording mode of the CD-ROM is mode 1 are as shown in FIG. That is, a sync (synchronization) pattern is arranged at the head of a sector, and a header including a sector number and a track number is arranged subsequently. And after this header is 2K
Bytes of user data are provided, and the last is auxiliary data including codes for error detection and error correction of the user data.

In the case of this example, the above-mentioned moving image data and other data are recorded in the user data area of the sector. The first 32 bytes of the 2K-byte user data are used as identification information ID. The identification information ID is information indicating what data is recorded in the user data area and indicating how many sectors of the data having the same content continue. It is needless to say that other information can be included.

As will be described later, the contents of the data indicated by the identification information ID include the user data of the sector, the image data of the moving image, the information of the color conversion table and the screen table scr, and the image of the background still image. Data, resident character data, game condition change data, etc. are prepared.

By the way, the above-mentioned data relating to the image for one frame includes the compressed image data relating to the pixel of each character in the 2-bit mode and the 1-bit mode, and the eight data of the frame other than the image data. Color conversion table COL shown in FIG. 7 for palette
(0) to COL (7) and the screen table scr shown in FIG.

In the case of this example, the compressed image data for one frame to be recorded has, as shown in FIG. 12, a mode in which the number of characters N in the 2-bit mode and the number M of characters in the 1-bit mode are provided at the beginning. Number information and compressed image data dat2 (n) (n
= 0,1,2... N-1) and M pieces of compressed image data dat1 (m) (m = 0,1,2... M-1) of 1-bit mode characters. As described above, a single-color character has its color information registered in the screen table scr.
It is not recorded as pixel data.

The information for one character is shown in FIG.
As shown on the lower side of the figure, the information includes color number table information and index number data for 64 pixels. As shown in FIG. 11, the index number data corresponding to each pixel is 2 bits in the 2-bit mode and 1 bit in the 1-bit mode. In this case, since the number N of characters in the 2-bit mode and the number M of characters in the 1-bit mode change according to the content of the pixel, the data length of data for one frame of character pixels is variable.

In this example, the number of characters in each mode is recorded as mode number information. Instead of this mode number information, the last character of the 2-bit mode and the first character of the 1-bit mode are recorded. In the meantime, mode delimiter information of a bit pattern that does not occur as character data may be recorded.

In the case of this example, the data amount of the moving picture compressed as described above is as follows for one frame.

Since there are 8 pallets per frame,
As a color conversion table, 16 (color) × 8 (palette) × 2 (byte) = 256 in total
(Bytes). Further, since the screen table scr has 2 bytes per character, 768 × 2 (bytes) = 1536 (bytes).

Therefore, the total data amount of the color conversion table and the screen table scr is 2 Kbytes or less, and can be accommodated in one sector.

Also, since the moving image data has four types of color numbers represented by 4 bits in a character in the 2-bit mode, the color number table is 4 (bits) × 4 = 16 (bits). = 2 (bytes). Since the index number data is 2 bits, 2 (bits) × 64 = 128 (bits) = 16 (bytes). Therefore, one of the two-bit mode characters
The data amount per character is 18 bytes.

Since the color number of the character in the 1-bit mode may be two colors, the color number table is as follows: 4 (bits) × 2 = 8 (bits) = 1 (byte). Also, since the index number data is 1 bit, 1 (bit) × 64 = 64 (bit) = 8 (byte). Therefore, one of the 1-bit mode characters
The data amount per character is 9 bytes.

Since the pixel data of a single color character is not transmitted, the compression ratio of one frame of image data is determined by the ratio of the number of 2-bit mode and 1-bit mode characters in one frame to the number of single-color characters. Is determined by For example, in the case of 2-bit mode: 1-bit mode: single color = 3: 3: 2 = 288: 288: 192, character pixel data 2-bit mode 288 × 18 = 5184 bytes 1-bit mode 288 × 9 = 2592 bytes Total 7776 Bytes, which is less than 8K bytes, so it fits within 4 sectors.

From the above, in the case of this example, FIG.
As shown in (1), data relating to a moving image can be recorded as 5 sectors for each frame. That is, 5
Character data in the 2-bit mode and the 1-bit mode (mode number information is included in the first sector) are recorded as user data of the first four sectors in the sector. The data of the screen table scr and the color conversion table are recorded in the fifth sector indicated by hatching in FIG. 14A.

As the 32-byte identification information ID of the user data area of each sector, the first four sectors include information indicating moving image data,
Information on the number of sectors that follow (4 in the case of the first sector) is recorded. In the last fifth sector, information indicating that the data is data of the screen table scr and the color conversion table, and information of the number of sectors following the data (in this case, 1) are recorded.

In this manner, data relating to a moving image for one frame is repeatedly recorded every five sectors. In this case, the sum of the data for one frame including the screen table scr, the color conversion table, and the image data is:
9568 bytes, and the CD-ROM transmission rate is 1
Considering that it is 50 Kbytes (75 sectors) / sec, in this example, a moving image of 15 frames (frames) / sec can be recorded or reproduced.

In this example, the CD-ROM contains
Resident characters and other data input from the terminal 39 are also recorded in the user data area of this sector. Furthermore, in the game machine of this example, as described later, a background image of a moving image can be separately created as a still image in a decoder, and a combined image of both can be displayed on a CRT display. ing. For this reason, still image data is also recorded in the CD-ROM. In this case, as still image data, 16-bit data per pixel is compressed to 4 bits and recorded. Resident characters are similarly compressed to 4 bits and recorded.

As shown by hatching in FIG. 14B, these data are inserted and recorded between one frame of data related to the above-described moving image. Then, as the first identification information ID of the user data area of each sector, the data content and the number of sectors following the data content are recorded.

Note that, in addition to the above-described compressed image information, a program for decoding the compressed image information and a game program are recorded on the CD-ROM. Further, audio information is also appropriately recorded. As a decoding program, a 2-bit mode decoding program and a 1-bit mode decoding program are recorded. Also, a character rearranging program is recorded. These program data are recorded separately from the data as described above, and at the time of decoding, prior to reproduction of a moving image, etc.
It is designed to be able to read all at once. Note that these program data may be recorded in the middle of the moving image data as data other than the image data of the moving image in the above example.

As described above, according to the data compression method described above, the image is hierarchically divided into small areas in units of one frame, and the vector quantization is performed on the image data of each layer. Data compression ratio can be increased.

In this case, one group (palette) is formed by grouping characters having similar colors, a plurality of groups are formed for one screen, and image data is divided into pallets (groups). I have. And
Since the vector quantization processing is performed in the palette including the image parts of similar colors, the quantization error is reduced.

Further, at the time of decoding, since the decoding process can be performed only by referring to the table, the configuration of the decoder is simplified. Further, since a large-capacity buffer memory is not required, a general-purpose DSP having a limited capacity of the built-in RAM can be used as a decoder, and the cost of the decoder can be reduced.

Further, since the compression processing is performed without using the frame correlation, even if an error occurs during decoding, the error is completed within one frame and does not affect subsequent frames.

Further, since the decoder circuit can be provided at low cost and a CD-ROM can be used as a recording medium, it is effective when applied to software of a computer game machine.

In the above example, the vector quantization by the vector quantization means 24 is performed by S / N in each frame.
Is maintained constant in all frames so that the maximum value Emax of the quantization error in the character is kept constant. For this reason, the data size after quantization changes according to the information amount of a frame (complexity of image content).

However, by quantizing each character as follows, the data amount (data transmission rate) per frame can be made constant.

That is, first, the threshold value Eθ of the distance that brings together pixels of similar colors in the character.
Is set, and vector quantization is performed for each character based on the threshold value. That is, for the image data in each character, the quantization error
Vector quantization is performed until just before exceeding θ. By this quantization, data of a character having a large color change is subjected to data compression so as to have four colors. Further, in a character having a flat color change, the number of colors is reduced, and some characters become three colors, two colors or one color.

When the vector quantization process is completed for all the characters in one frame, the maximum value Emax of the quantization error in all the characters in one frame is calculated. Next, the number N of characters in the 2-bit mode, the number M of characters in the 1-bit mode, and the number L of single-color characters in one frame are counted. Next, these numerical values N, M, L
To calculate the image data amount per frame. The calculation of this pixel data amount is as follows.

Data amount of one frame = N.times.18 (bytes) + M.times.9 (bytes) + L.times.0 As a result, whether the data amount of one frame is equal to or smaller than a predetermined value, and therefore, the compression rate is equal to a predetermined value It is determined whether or not the threshold value E is reached.
is set to the maximum value Emax of the quantization error, and the above-described vector quantization processing is repeated.

As described above, the vector quantization is repeated by changing the threshold value Eθ until the data amount per frame reaches the predetermined data amount. In such a case, the S / N differs for each frame, but the amount of transmission data is constant. That is, in the case of a moving image to be described later, the number of frames (frames) per second can be constant.

In the above example, the data of the character having one color is stored in the screen table scr.
, And only the color data is transmitted, and the data on a pixel-by-pixel basis are not transmitted.

The processing unit for dividing the pallet is 1
Instead of frames, a plurality of frames may be used to perform three-dimensional palette division.

[Explanation of Decoding Device as One Embodiment of the Invention] Next, the CD-R
A case of a game machine as an example when the present invention is applied to a device for decoding image data recorded in an OM will be described.

FIG. 1 shows an example in which the present invention is applied to a game machine using a microcomputer, wherein 1 is the game machine body, 4 is a sub-processing section, and 5 is a CD.
ROM, 6 is a program cartridge, and 7 is a main processing unit for audio data.

The game machine body 1 is constituted by a microcomputer, 11 is its CPU, 12
Denotes a DMAC (DMA controller), 13 denotes a RAM for a work area, 14 denotes a PPU (Picture Processing Unit), and 15 denotes a video RAM.

Then, the game machine main body 1 includes the first and second game machines.
2 bus configuration including the system buses 18 and 19 of FIG. The two system buses share a data bus, but have an address bus that is a first system bus and a second system bus.
System bus. And DMAC
12 allows DMA transfer only between the first and second system buses 18 and 19.

The first system bus 18 has a CPU 1
1, the DMAC 12 and the RAM 13 are connected. Also,
The DMAC 12 and the PPU are connected to the second system bus 19.
14 is connected and a video RAM is
15 and the CRT display 6 are connected. The sub-processing unit 4 and the main processing unit 7 for audio data are connected to the second system bus 19.

In this case, the CPU 11 and the second system bus 19 are connected via the port 16 and
The device connected to the first and second system buses 19 can be accessed via the port 16.

In this case, the video RAM 15 is divided into a plurality of memory areas M1 to M4, for example, as shown in FIG. In the case of this example, the memory area M1 is an area for reproducing a moving image, the memory area M
2 is an area for a background still image, and the memory area M3 is an area for resident characters and the like. Each of these memory areas M1 to M3 has a screen area of two frames (two screens), and image data of one of the screen areas is read by the PPU 14 in synchronization with vertical and horizontal scanning of the CRT display 8. The image is displayed on the display 8 as an image, and while this display is being performed, the image data of the image to be displayed next is written in the other screen area.

The memory area M4 is a work area of the PPU 14, and is used as a screen table scr, a color conversion table, and other data areas.

Further, in the audio data main processing section 7 of the game machine main body 1, reference numeral 71 denotes its APU (audio processing unit), 72 denotes a D / A converter,
Reference numeral 3 denotes an audio output terminal. The APU 71 is connected to the bus 19 and connected to the D / A converter 72. When audio data and a decoding program for the audio data are loaded into the APU 71, the audio data is decoded into a digital audio signal, and the digital audio signal is D / A converted into an analog audio signal by the converter 72, and then output to the output terminal. 73 is output.

The sub-processing section 4 has a CD player to enable the use of the CD-ROM 5, and
1 is the CD player, 42 is the DSP, 43 is the CD-R
OM decoder, 44 is RAM for the work area, 4
5 is a controller. The audio data and the image data are recorded on the CD-ROM 5, and the audio data and the image data, especially the image data, are recorded by being compressed as the image data by the above-described method.

The DSP 42 corrects errors in the reproduction signal of the player 41 and separates user data such as image data and control data such as track numbers from the reproduction signal.
5 is the control data from the DSP 42 and the CPU 11
Control the player 41 based on the instruction data from
This is for reproducing target data. Also,
The decoder 43 outputs the playback signal of the player 41 to a CD-RO
In the case of an M5 reproduction signal, this is for performing processing such as error correction for the CD-ROM.

Further, in the sub-processing unit 4, 50 is DS
In P, which is a general-purpose DSP, the sub-processing unit 4 processes image data. Although the sub-processing unit 4 is integrated with the game machine main body 1 in this example, it may be in the form of an adapter with respect to the game machine main body 1. Although not shown, the DSP 50 includes a program RAM and a buffer RAM (which can be constituted by one RAM).

The program cartridge 6 is used by being inserted into the slot 2 of the game machine body 1 when using the game machine. When the CD-ROM 5 is not used, a program cartridge for general game software is inserted into the program cartridge 6.
When using, a special one is inserted.

The cartridge 6 has a ROM 61
And a RAM 62, and in the case of a cartridge for a CD-ROM 5, the ROM 61 includes a CD-ROM 5
A program for a so-called initialization process for causing the game machine main body 1 to take in the recorded data and execute the game is written. The RAM 62 is used, for example, when temporarily interrupting the game in the middle of the game, for holding various data relating to the state at that time until the game is restarted, and is backed up by the battery 63.

When the cartridge 6 is inserted into the slot 2 of the game machine main body 1, the ROM 61 and the RAM 62 are connected to the bus 18 via a connector (not shown).

Then, the program in the ROM 61 of the cartridge 6 is executed by the CPU 11 and the CD-ROM 5
Is taken into the RAM 13 of the game machine body 1, and the decoding process of each user data is performed based on the identification information ID in the user data of each sector. Thus, the moving image is displayed, and the still image and the resident character of the background are displayed without stopping the moving image.

That is, the CD-player 41
When the data is reproduced from the ROM 5, the reproduced data is sequentially supplied from the player 41 to the DSP 42 and the decoder 43 to perform processing such as error correction, and the error-corrected data is transmitted from the decoder 43 to the DMAC 12 by the DMAC 12. DMA in the first buffer area of the RAM 13
Will be transferred.

Next, the CPU 11 checks the identification information ID of each sector of the data fetched into the RAM 13. According to the check result, the CPU 11
Executes a decoding procedure corresponding to the reproduction data of the content indicated by each ID.

The decoding program and the game program described above are loaded into the RAM 13 from the CD-ROM 5 prior to reproduction of image data and the like.

[Decoding of Moving Image Data] CP
As a result of checking the identification information ID in U11, when it is determined that the contents of the user data of the sector are image data of a moving image, the moving image is decoded and displayed as follows.

That is, the decoding process is performed on the data of 5 sectors including the compressed image data of one frame as follows. The procedure for decoding the moving image data basically includes the following three steps.

A. For each character, referring to the color number table, 2-bit or 1-bit index number data is converted into color conversion tables COL (0) to COL (7).
B. Step of referencing the first-order table for conversion into 4-bit color number data B. For each pixel of the character of each palette, refer to the color conversion table of the palette, and convert the data of the color number decoded in section A into actual color data. Step of rearranging the sorted characters, that is, B
Rearranging the pixel data decoded by the term to the original character position DSP 50 performs the step of the term A among the steps of the terms A to C, and the steps of the terms B and C
U14 does.

[1. Step A) First, the DSP 50 decodes the user data of four sectors including the compressed image data of one frame into data of the color number as follows, and converts the decoded data to the video RA.
The procedure up to writing in the memory area M1 of M15 will be described. That is, (1) A program for decoding a 2-bit mode character is loaded from the RAM 13 to the DSP 50.

(2) Of the 2-bit mode character data of the image data DMA-transferred to the first buffer area of the RAM 13, eight characters from the beginning of the character data are DMA-transferred to the DSP 50 by the DMAC 12.

(3) In the DSP 50, the step of item A is executed by the program of (1), and the index number data transferred by DMA is converted into color number data (FIG. 11A) by the color number table. By this conversion, the index number data (=
18 bytes x 8) is 4 bits x 8 pixels x 8 pixels (= 25
(6 bytes) color number data.

(4) The decoded color number corresponds to the DM
AC12 stores D in the second buffer area of RAM13.
MA transfer is performed.

(5) Thereafter, the processing of (2) to (4) is repeated, and all of the index number data of the character in the 2-bit mode are decoded into color numbers and DMA-transferred to the second buffer area of the RAM 13. You.

(6) The data of all the color numbers in the 2-bit mode DMA-transferred to the second buffer area of the RAM 13 is transferred to the video R through the PPU 14 by the DMAC 12 during the vertical blanking period of the CRT display 8.
The data is DMA-transferred to the AM 15 and written to the memory area M1.

(7) When the processing up to (6) is completed, a program for decoding a 1-bit mode character is loaded from the RAM 13 to the DSP 50.

(8) Of the character data in the 1-bit mode of the image data DMA-transferred to the first buffer area of the RAM 13, eight characters from the head thereof are DMA-transferred to the DSP 50 by the DMAC 12.

(9) In the DSP 50, the step of item A is executed by the program of (7), and the index number data transferred by DMA is converted into color number data (FIG. 11B) by a color number table. By this conversion, the index number data (=
9 bytes x 8 pixels are 4 bits x 8 pixels x 8 pixels (= 25
(6 bytes) color number data.

(10) The decoded color number is
AC12 stores D in the second buffer area of RAM13.
MA transfer is performed.

(11) Thereafter, the processing of (8) to (10) is repeated, and all of the index number data of the 1-bit mode character is decoded into color number data, and
13 is DMA-transferred to the second buffer area.

(12) The data of all the color numbers in the 1-bit mode DMA-transferred to the second buffer area of the RAM 13 is transferred to the video R through the PPU 14 by the DMAC 12 during the vertical blanking period of the CRT display 8.
The data is DMA-transferred to the AM 15 and written to the memory area M1.

The DMA transfer of the 2-bit mode color number in (6) can be performed immediately before (12) (between (12) and (11)).

[2. Steps of Items B and C] (13) When the processing up to the step (12) is completed, the processing of the fifth sector of the image data of one frame is started. That is, the CPU 11 detects that the fifth sector is a data sector of the screen table scr and the color conversion table based on the identification information ID. Therefore, CPU1
1 transfers the data of the screen table scr and the color conversion table DMA-transferred to the first buffer area of the RAM 13 to the DMAC without passing through the DSP 50.
12 to the video RAM 15 through the PPU 14
A transfer. In this case, these screen tables sc
The data of r and the color conversion table are written to the memory area M4 of the video RAM 15.

(14) When the above transfer processing is performed, the PPU
14 executes the steps of the above-mentioned B and C terms in real time. That is, by referring to the color conversion table COL (j), the data of the color number of the memory area M1 processed by (2) to (5) and (8) to (11) is converted into the pixel data of the actual color. , And by referring to the screen table scr, the pixel data of each character is written to the address corresponding to the original character position.

(15) When the pixel data for one frame is written to one screen area of the memory area M1 of the video RAM 15 as described above, the display area of the video RAM 15 is switched to the screen area, and the pixel data is written. The activated area is activated, and its screen is displayed on the display 8.

(16) The process returns to (1), and thereafter, the processes of (1) to (16) are repeated for each frame.

The image data reproduced from the CD-ROM 5 is stored in the RAM 13 and the DSP 50 as described above.
And the PPU 14 are successively sent to the video RAM 15 while being processed in a pipeline processing. Therefore, an image based on the image data of the CD-ROM 5 is displayed on the display 8 as a moving image. This moving image display can be performed at a rate of 15 frames / second as described above.

[Other Data Processing] As described above,
Data sectors of the screen table scr and the color conversion table are written to the video RAM 15 as sectors other than the moving image in a different process from the sectors of the moving image data. Similarly, still image data, resident character data, and other data sectors are also processed by the video RAM 13 by a process different from that for moving images based on the identification information ID.
Written to other memory.

That is, the reproduced data from the CD-ROM 5 is as shown in FIG. 3, and as shown by hatching in the figure, one frame (5 sectors) of moving image data is shown.
If a sector of other data is inserted between
At the beginning of the inserted other data, the start pointer PS for the processing of the other data is raised, the processing of the other data is started, and the decoding processing of the moving image is stopped, and the moving image is displayed on the previous screen. Will remain.

As for the other data, the identification information ID of the sector is checked by the CPU 11, the content of the other data and the number of continuous sectors are detected from the identification information ID, and the other data is stored in an appropriate memory area. Alternatively, the data is written into another memory and processed according to the content of other data. When the sector of the other data ends, the end pointer PE rises, and the processing of the other data ends. If the subsequent sector is a moving image sector, the decoding process of the moving image data is restarted. Therefore, the moving image continues from the previous screen. In this case, 5 sectors = 1/15
Since the second data is a short period and the other data insertion section is a short period, the moving image visually looks almost continuous.

When the other data is, for example, a still image, the data is processed as follows and written into the memory area M2 of the video RAM 15.

(1) First, the CPU 11 checks the identification information ID of the user data of the sector, determines that the sector is a still image, and shows the sector where the still image continues, for example, as shown in FIG. As described above, when it is determined that there are five sectors, the start pointer PS and the end pointer PE indicate the interval between these five sectors, and the process shifts from the moving image decoding process to the temporary and still image processing. Then, the instruction is provided from the CPU 11 to the PPU 14.

(2) The image data of the still image DMA-transferred to the RAM 13 is transferred to the memory area M2 of the video RAM 15 via the PPU 14 without passing through the DSP 50.
DMA transfer.

(3) The PPU 14 performs a decoding process of returning the image data of the still image from 4 bits to the original 16-bit data, and rewrites the decoded still image data to the memory area M2. This data decompression processing is executed when a general program cartridge is used as a program cartridge, and the program is prepared in the PPU 14 in advance.

The still image decoded as described above is combined with a moving image and used as a background of the moving image.

As described above, since the identification information ID indicating the data content and the number of sectors in which the data having the same content continues is recorded for each sector in the CD-ROM, the decoding device monitors the identification information ID. By doing so, decoding processing can be performed for each sector. Therefore, other data sectors can be inserted and recorded between image data of a moving image in which one frame includes a plurality of sectors, and a plurality of sectors can be inserted in units of sectors. It is possible to insert a large amount into moving image data.

As a moving image, if the previous screen is held during the period of the other data, the moving image follows the period of the other data, so that the moving image can be reproduced without stopping visually.

Further, according to the example shown in the figure, all data flows are managed by the CPU 11 so that the CD-ROM
5, since the CPU 11 absorbs the asynchronous operation between the reading of the image data 5 and the processing of the CPU 11,
, The image data can be read continuously.
Moreover, the configuration for this is simple as is clear from FIG.

The DSP 50 performs the primary decoding and the PPU 14 performs the secondary decoding on the image data of the moving image that has been data-compressed. Can be used, and the cost can be reduced.

Further, since the decoding of the image data which has been subjected to data compression is performed separately by the DSP 50 and the PPU 14, the image data can be decoded at a sufficient speed, and a moving image with a sufficient motion is displayed. be able to.

The RAM 13, DSP 50, PP
Since the DMAC 12 transfers data to and from the U 14, it does not impose a load on the CPU 11. In addition, DSP50
Since the CPU 11 is vacant while is performing decoding, it can instruct processing of other data as described above. As other data, audio data can be inserted. In that case, the user data of the sector of the audio data is transferred to the RAM of the APU 71.

As a recording medium, not only a CD-ROM but also a tape or the like can be used.

[0129]

As described above, according to the present invention, the processing apparatus is provided with two system buses, which are capable of performing DMA transfer with the two system buses. Since the first buffer memory is connected and the sub-processor for reproducing data from the recording medium is connected to the other system bus, the data from the sub-processor can be DMA-transferred to the first buffer memory. . The sub-processing unit is a CD-ROM player and a CD-RO
In the case where the first buffer memory is constituted by an M decoder, a large amount of data from the CD-ROM decoder is stored in the first buffer memory.
MA transfer is possible.

Therefore, there is an advantage that the transfer of data from the sub-processing unit and the processing of the CPU can be performed asynchronously, and the configuration of the apparatus is simplified.

The CPU manages all the data flows, so that, for example, the C
Asynchronism between the reading of the image data from the D-ROM and the processing of the CPU can be absorbed by the CPU itself, and therefore, the image data can be continuously read from the CD-ROM.

When a data processor is connected to the other system bus, data can be DMA-transferred from the first buffer memory to the second buffer memory via the data processor. By performing predetermined data processing such as image data processing, data from a recording medium such as a CD-ROM can be processed in real time. Therefore, for example, a moving image having a sufficient number of frames can be reproduced in real time.

[Brief description of the drawings]

FIG. 1 is a block diagram of an embodiment of a decoding device according to the present invention.

FIG. 2 is a diagram for explaining a divided storage area of a memory.

FIG. 3 is a diagram for explaining other data inserted into moving image data.

FIG. 4 is a partial block diagram of an example of an encoding apparatus that implements an embodiment of a method of compressing image data recorded on a recording medium according to the present invention.

FIG. 5 is a block diagram of a remaining part of an example of an encoding device that implements an embodiment of a method of compressing image data recorded on a recording medium according to the present invention.

FIG. 6 is a diagram for explaining an example of a method of dividing image data according to an embodiment of the image data compression method of the example of FIGS. 4 and 5;

FIG. 7 is a diagram for explaining a table used in an embodiment of the image data compression method in the examples of FIGS. 4 and 5;

8 is a diagram for explaining an example of compressed data according to an embodiment of the image data compression method of the example of FIGS. 4 and 5. FIG.

FIG. 9 is a diagram for explaining an embodiment of the image data compression method in the examples of FIGS. 4 and 5;

FIG. 10 is a diagram for explaining an example of a table used in an embodiment of the image data compression method in the examples of FIGS. 4 and 5;

FIG. 11 is a diagram for explaining an example of compressed image data recorded on a recording medium according to the present invention.

FIG. 12 is a diagram showing an example of a format of data recorded on a recording medium according to the present invention.

FIG. 13 is a diagram for explaining a sector format of a CD and identification information ID for each sector recorded on a recording medium according to the present invention.

FIG. 14 is a diagram showing an example of reproduction data from an example of a recording medium according to the present invention. .

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Game machine main body 5 CD-ROM 6 Program cartridge 8 CRT display 11 CPU 12 DMA controller 13 RAM 14 PPU (Picture Processing Unit) 15 Video RAM 22 Character dividing means 24 First-stage vector quantization means 25 Pallet dividing means 27 Vector quantization means of second stage 38 Recording processing means 39 Input terminal of other data 41 CD-ROM player 50 General-purpose DSP

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁶ , DB name) G06T 1/00-1/20 G06F 13/20-13/34 G06F 13/38-13/40 H04N 7 / 24-7/62

Claims (2)

(57) [Claims] And 1. A processing apparatus main body, and reproducing data from a recording medium on which compressed image data is recorded, the decoding
Consists of a sub-processing unit to obtain compressed image data, the processing apparatus main body, a CPU, along with it can be accessed from the CPU, between each other
Two and a system bus capable DMA transfer in a first buffer memory connected to one of the two system bus, which is connected to the other of the two system bus Rutotomoni, a data processor and a data processing unit and a second buffer memory, the sub-processing unit, which is connected to two of the other system bus of the system bus, wherein the compression obtained by decoding by the sub-processing unit Image
Chromatography data is DMA-transferred to the first buffer memory, the compressed image data from the first buffer memory
Is stored in the second buffer memory of the data processing unit.
And MA transfer, the pressure in said data processor
Data <br/> data processing apparatus characterized by extending decode a reduced image data. Wherein said auxiliary processing unit includes a CD-ROM player, a CD-ROM decoding unit, wherein the CD-ROM decoding unit of the sub-processing unit, the two system bus of the processing apparatus main body The other of the above
2. The data processing device according to claim 1, wherein the data processing device is connected to a system bus.