JPH1188700A - Coding method of color image signal, and decoding method and color image processor thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a coding method in which an image signal of the pallet color system is compressed with a high efficiency and to provide an image- processing unit where high-speed signal processing is conducted and is simplified through the coding method. SOLUTION: In the coding method for a color image signal, where a color signal of each pixel of a color image is expressed by a pallet color registered on a pallet table, an image with one screen portion is divided into blocks each of which includes a prescribed number of (m×n) pixels (s102), and the number of kinds of colors used in each block is limited to a prescribed number smaller than a fixed number (m×n), depending on an incidence state of pallet colors of each pixel in each block and a block index signal (XB) for referencing a pallet table is set to each limited color kind (s103). When plural block index signals are in existence, a selection signal (s) for making any of the signals correspond to each pixel is set (s104), and the image signal of each pixel is coded by the selection signal and the block index signal and then compressed (s105).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for a color image signal, and more particularly to an encoding and decoding processing system using a palette color system.

[0002]

2. Description of the Related Art In order to display a color image, R
A combination of (red) G (green) B (blue) or C (cyan) M (magenta) Y (yellow), or a signal such as Yuv or Yab obtained by a conversion process is used. For example, if an 8-bit binary number is assigned to each of the three colors to represent the color of one pixel, 24 pixels per pixel are obtained.
It becomes a bit multi-valued signal, and the color type is (2 ⁸ ) ³ ≒ 160
It will be 10,000. Such a multi-level signal is used to represent a natural image or the like input by, for example, a scanner.

On the other hand, when the types of appearing colors are limited, such as a character graphic image created on a personal computer screen,
By predetermining a combination of three color signals and indicating which combination of color type each pixel selects,
A color image signal can be configured. For example, when 256 or less colors appear on one screen, one color can be selected from 256 types by assigning an 8-bit signal to one pixel. Specifically, 256 kinds of color signals (8 bits for each of RGB) are set in a memory in advance, and read out using a memory address generated from an index signal (8 bits) of each pixel, so that each pixel is read out. Outputs three color signals.

[0004] This color signal setting method is called a palette color method, and a palette table storing palette colors is prepared. Although the color type of the palette color system is limited, it is possible to reproduce a color signal with high bit precision (24 bits per pixel in the above example) and to compress the data capacity (one third in the above example). There is.

JP-A-6-284269 (Cited Example 1)
There is a proposal concerning compression of an image signal by a palette color system. According to this, when the color type represented by the palette color is smaller than the size of the table prepared in advance, the size of the table is reduced according to the appearing color type. For example, when the size of the table prepared in advance is 256 and the number of colors appearing on the screen is eight, the size of the table is reduced to eight. As a result, a signal of 8 bits per pixel required for table search can be compressed to 3 bits.

The images created by a personal computer or the like include a natural image of 8 bits for each color of RGB input from a scanner or the like, and an image of a character figure using a palette color. Since the configurations of the two image data are completely different, the input device, the storage device, the display device, and the like are used in the configuration of the image processing device.

[0007]

In recent image creation using a personal computer or the like, it is often used to edit a natural image such as a photograph or a picture taken from a camera or a scanner by partially pasting the image. As described above, when a screen is created by partially capturing a natural image, the method of the cited example 1 has a large number of color types of a part of the natural image, so that it is difficult to reduce a palette table relating to the entire screen. There is a problem that the effect of image compression cannot be obtained.

By the way, the present inventors have disclosed Japanese Patent Publication No. 6-768.
As proposed in No. 8 (Cited Example 2), a “color image information encoding processing method” for natural images has been realized. Here, image compression is performed on a small area of pixels having multi-level signals (for example, 4 × 4 pixels) by limiting the number of appearing colors to several types. According to this method, although a slight signal deterioration occurs, there is almost no visual deterioration in image quality.

Based on this coding method, the present invention overcomes the problems of the conventional palette color system and is effective when applied to a palette color image partially including a natural image. Reached.

SUMMARY OF THE INVENTION An object of the present invention is to provide an encoding and decoding method capable of efficiently compressing an image signal represented by a palette color and performing a high-speed decoding process, and simplifying an apparatus configuration by applying the encoding and decoding methods. Another object of the present invention is to provide a color image processing apparatus with improved functions.

[0011]

An object of the present invention is to provide a method of encoding a color image signal in which a color signal of each pixel of a color image is represented by a palette color registered in a palette table. For each block divided so as to include (m × n) pixels, the number of color types used in accordance with the appearance of each pixel in the block of the palette color is smaller than the predetermined number (m × n). A block index signal (XB) for referencing the palette table for each color type to be used, and a selection signal for associating one of the block index signals with the color type of each pixel when there are a plurality of block index signals. This is achieved by creating compressed data that is more encoded with (S).

The color type to be used may be limited to a fixed value preset in the order of the appearance frequency of the palette color in the block, or the appearance frequency may be equal to or greater than a preset lower limit and be equal to the predetermined number (m). Xn) Depends on the following variable values. Alternatively, a component of a color signal of a pixel in a block (for example, RG
B), the color signal of each pixel is grouped using the intermediate value as a threshold, and then the maximum amplitude of the color signal component of the pixel in the group is determined, and the intermediate value is set as a threshold. The number of fixed groups obtained by repeating the process of grouping the color signals of the pixels a predetermined number of times, or
When the maximum amplitude is equal to or larger than the set amplitude value, the grouping is advanced. When the maximum amplitude is smaller than the set amplitude value, the grouping is stopped, and the number of variable groups obtained when the division ends is determined.

[0013] The compressed data is provided with a color type signal (C) indicating the number of color types used in the block, and the compressed data is structured for each block or for each of a plurality of blocks. Further, an identical color signal indicating whether or not the adjacent block and the block index signal are the same is added, and the block index signal is added only to the non-identical blocks. The same color signal can use one of the color type signals (C).

Further, the object of the present invention is to provide a method for decoding a color image signal in which compressed data of a color image signal is expanded by a palette color method, wherein the compressed data is stored in the palette table for each of a predetermined number of color types used in the block. Block index signal (XB) for referring to
And the selection signal (S) for associating the color type of each pixel with the block index signal, the compressed data is read out in block units, and the block index corresponding to each pixel is read from the selection signal (S) in pixel order. This is achieved by selecting a signal, using the block index signal to search the pallet table, and decoding to a corresponding pallet color.

In the decoding, the compressed data is sequentially read out in block units from the first block to the last block arranged in the line direction of the screen, and the compressed data of the same line in each block is continuously read from the first block to the last block. After repeating the decoding process for n lines, the process shifts from the first block arranged in the line direction at the next stage of the screen to the decoding process of the last block.

The second block is read while decoding the read compressed data of the i-th line (i ≦ n) of the first block, and the second block is read after the completion of the decoding of the i-line of the first block. A replacement process for starting decoding of the i-line of the block is continuously repeated until the last block to output a color image signal of the i-line.

A drawing apparatus to which the encoding method of the present invention is applied includes an image signal storage device for storing a color image signal created or input by a palette color system, and a compressed data storage for storing compressed data of the color image signal. A device, a palette table for storing an index to a predetermined number of palette colors and storing the same, an encoding unit for encoding the color image signal to create the compressed data, and an editing unit for controlling each unit. The encoding means includes: for each block obtained by dividing an image for one screen including a fixed number (m × n) of pixels, according to the appearance of a color type based on a color image signal of each pixel; A block index signal (XB) for referencing the palette table for each color type limited to a number smaller than the predetermined number (m × n) And setting a selection signal (S) for associating the color type of each pixel with the block index signal, and creating the compressed data based on the selection signal and the block index signal.

Further, storage means for storing the block index signal and the selection signal read out for each block from the compressed data storage means, and selection means for selecting one of the stored block index signals using the stored selection signal. A decompression device including decoding means for extracting a corresponding color signal from the palette table using the selected block index signal.

Further, for each block obtained by dividing the natural image including a fixed number (m × n) of pixels, a natural image is selected from a block index signal representing a plurality of approximate colors approximating the inside of the block and a selection signal of each pixel. When compressed data of an image is created, a gamma table for searching for the approximate color is further provided, and the decompression device can be shared according to a palette image or a natural image.

The operation of the color image signal encoding method of the present invention will be described with reference to FIG. First, a color image signal of one screen pixel is input or created and stored in the image signal memory (s101). Here, the content of the color image signal is a palette color according to a palette table. next,
An image for one screen is divided into a plurality of blocks each consisting of a fixed number (m × n) of pixels (s102), and the following encoding process is performed for each block.

That is, according to the appearance state of the palette color of each pixel, the color type used in the block is set to a predetermined number smaller than the predetermined number (m × n), and the palette table is referred to for each color type. The block index signal (XB) is set (s103). For example, the types of pallet colors in a block are measured, and a predetermined number (fixed) or a predetermined number (variable) of a certain frequency or more in order of appearance frequency
And the corresponding index of the pallet table is set in the block index signal (XB) assigned to each color type.

Next, a selection signal (S) for associating its own color with the block index signal (XB) is set for each pixel in the block (s104). For example, each pixel can select the block index signal (XB) that is the same as or closest to its own palette color (distance in the color space). Next, a color image signal of each pixel is represented by a selection signal (S) and a block index signal (XB), and a compressed signal is created in a predetermined format for each block (s10).
5). Then, the above encoding process is repeated until the end of all blocks (s106).

[0023]

Embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 2 shows the configuration of an image processing apparatus according to one embodiment of the present invention. An input device 10 for inputting a target image signal, an image data storage device 11 for storing image signals for at least one block, usually one to several screens, an encoding device 12 for generating compressed data, A compressed data storage device 13 for storing, a decoding device 14 for reproducing image data from the compressed data, an editing device 15 for setting parameters and the like for the encoding device 12 and the decoding device 14, and displaying or printing a decoded image signal. It comprises an output device 16 and a pallet table 100. The storage device of the pallet table 100 is used individually or in common with the compressed data storage device 13 or the like.

The editing device 15 inputs or generates an image to be encoded with the intervention of an operator, and sets the characteristics of the image to be encoded, the target compression ratio, the target image quality, the setting of various parameters, the setting of additional information, and the like. And controls the encoding device 11 to create and store compressed data. Further, an appropriate parameter is set for the processing speed of the decoding device 14 and the like, and the decoding process of the image signal is controlled. Further, the output device 16
To control the display or printing of the image signal. As described above, the procedure can be automatically advanced by the processing program installed in the editing device 15 or can be advanced based on the setting by the operator.

Although the editing device 15 may be configured as a dedicated device, in this embodiment, it is a general-purpose computer such as a PC or WS. In this case, the editing device 15 can also create an image signal. When a general-purpose computer system is used, a TV camera, a scanner, a disk drive, a communication unit, or the like can be used for the image signal input device 10. For the image signal storage device 11 and the compressed data storage device 13, a storage medium such as a magnetic disk, CD-ROM, or DVD can be used.

The decoding device 14 in the configuration of FIG.
Is not used (or is not provided), and the output device 16 is used.
There is also a system configuration in which compressed data is transmitted to another image processing apparatus by using the communication medium as a communication medium. At this time, the other image processing apparatus does not have the encoding apparatus 11 (or does not use the encoding apparatus 11), and reproduces the input compressed data by the decoding apparatus 14 and then displays an image on the output apparatus 16 for example. Becomes

The pallet table 100 has a set of combinations of color signals required for image display. The combination of color signals includes, for example, R (red), G (green), B (blue)
Or C (cyan) M (magenta) Y (yellow) or Y
Signals such as uv and Yab can be used. If each color signal is 8 bits, the color type of the pixel can be expressed using 24 bits obtained by combining three colors. Pallet table 10
0 is a pallet table reconstruction unit 110 as described later.
Thus, it is also possible to reduce the amount of data and perform reconfiguration.

FIG. 3 shows the data structure of the image signal and the palette table. (A) shows the data structure of the input image signal. An image signal Xij at a pixel position (i, j) on the screen is set by an index Pk for reading out the contents of the pallet table 100. (B) is a cutout of a data structure per block of an input image signal, and a block Bij is composed of m × n pixels having pixel ij as an end point.

FIG. 3C shows the data structure of the compressed data. The compressed data includes a block index signal XB indicating a color type c (c <m × n) smaller than m × n appearing in the block, and a block index signal X representing each pixel.
A selection signal Sij for selecting B. If the pixel position in the block is represented by (i, j), in order to reproduce the image signal from the compressed data, the block index signal XB is selected using the selection signal Sij of each pixel, and XB is stored in the palette table 100. And reads the contents of the table. Thus, the color signals RGB of the image signal Xij can be output.

For example, if the pallet table 100 has 256
The image signal of each pixel is an 8-bit index signal for selecting one of 256 types of palettes. When the block size is m = 4 and n = 4, the number of bits per block of the input image signal is the number of pixels (m × n = 4 ×
4 = 16) and the number of bits of the index signal (8 bits)
And 16 × 8 = 128 bits.

Here, if the color type of the block index signal XB is limited to four types and the selection signal S of each pixel is 2 bits necessary to select four types of XBc, the compressed data of each block is Since the block index signal XBc has 4 × 8 = 32 bits and the selection signal S has 16 × 2 = 32 bits, the number of bits is 64 bits in total.

When the number of colors of the block index signal XB is limited to three, since the number of bits is 3 × 8 = 24 bits, the total of the selection signals S is 56 bits. Also,
If the color type of the block index signal XB is limited to two types, 2 × 8 = 16 bits, so that the total is 32 bits. Further, when the color type of the block index signal XB is limited to one type, the same color signal is used in the block, so that the selection signal S of each pixel is not required, and only the block index signal XB has 8 bits. .

As described above, the input image is divided into blocks each having m pixels in the horizontal direction and n pixels in the vertical direction, and the number c of the block index signals XB indicating the types of colors appearing in the blocks is determined by the number m of pixels in the block. When the number is limited to less than × n, the amount of image data can be reduced.

In this embodiment, based on the color signal characteristics in the block, the number c of types of colors appearing in the block is represented by c = c
It can be set variably as in 1-4. For example, if the type c of the block index signal is 1, 2, 3,
4 can be selected. Then, in order to indicate the type c, a 2-bit color type signal C is added for each block. If the type of the block index signal XB is 4,
In the case of 2, 1 types, the number of bits of the compressed data per block is 66, 34, 10 bits by adding the 2-bit color type signal C to the above-described value.

As a result, the image quality can be maintained by increasing the number of color types for a block having a sharp change in color, while the compression rate can be increased by reducing the number of color types for a block having a small change in color. As described later, if the number of color types that appear corresponding to the local properties of the image is switched,
It is possible to improve the compression ratio and maintain the image quality.

Further, as a general property of the image signal,
The high correlation of the appearance color between adjacent pixels is mentioned. This is the same for adjacent blocks. Therefore, the block index signal itself can be omitted by indicating whether or not the block index signal between blocks is the same. To indicate whether they are the same, a one-bit same color signal K is added to each block. Alternatively, the above-described color type signal C can be shared. When the color type signal C is also used, for example, a 2-bit color type signal C is used, and 00 indicates the same between blocks (other than 00). , 4 are shown.

Note that it is essential to set an initial value at a certain timing depending on the configuration of the image signal and the scanning order. For example, since it is not possible to determine whether the first blocks in the scan order are the same or not, it is necessary to perform processing such as comparison with some initial value or avoiding the comparison.

FIG. 4 shows the structure of the compressed data generated by the above procedure. FIG. 7A shows the configuration of the same color signal K, color type signal C, block index signal XB, and selection signal S, and the number of bits per block. FIG. 2B shows the structure of the compressed data when the same color signal K is not used and the required number of bits. The block index signal XB is 8 bits each, the selection signal S is 32 bits per block, the color type signal C is 2 bits per block, and the same color signal K is 1 bit per block. Note that the correspondence between the selection signal S and the pixel position is such that when the starting end point of the block is i, j, S00 is the pixel (i,
j) and Smn are pixels (i + n, j + m).

When storing or transmitting compressed data in a memory, it is convenient to divide the data into multiples of 8 bits (1 byte) used in general data processing devices. FIG. 5 shows the arrangement of compressed data. If the same color signal K is not used for the four blocks A to D as in (a) or (b), as shown in (c),
If the color type signals for the blocks are put together at the beginning, 4 × 2 = 8
Since all bits are bits and all signals are a combination of 8 bits, it is easy to store the compressed data in the memory 13. As a result, since memory input / output and signal processing can be performed in units of 8 bits, software processing or the configuration of a hardware circuit is facilitated.

The compressed data storage device 13 can store some information in addition to the compressed data. For example, the header information includes a data title, a screen size, a compressed data capacity, and a pallet table. Further, by adding a processing program to the header information of the compressed data, it is also possible to execute signal processing such as reading and decoding of the compressed data.

Next, an embodiment of the encoding process of the present invention will be described. As shown in FIG. 1, in addition to the embodiment in which the types of colors appearing in a block are measured and the color types are set in descending order of the number of appearances, encoding processing according to the following procedure is possible.

FIG. 6 is a flowchart showing an encoding process according to another embodiment. Here, for the image signal X of one block, the color type appearing in the block is 4
This shows an encoding procedure limited to types. In order to facilitate understanding of this processing, FIG.
See

As initial settings, parameters that determine processing procedures such as the block size (for example, 4 × 4 pixels) and the upper limit of the color type are set from the editing device 15 (s201). Next, the image signal Xmn of the pixel in the block is read from the pallet table 100 and converted into a three-color signal (for example, an RGB signal) (s202). The converted color signals (RGB) of the pixels in the block are distributed in three-dimensional RGB color space coordinates as shown in FIG.

Next, an amplitude value is obtained from the maximum value and the minimum value of each of the RGB color signals. Then, a color type (R, G, or B) having the maximum signal amplitude is obtained (s203). In the color space, this corresponds to measuring the signal amplitude projected on the RGB coordinate axes.

Further, the intermediate value of the amplitude range of the color signal having the maximum amplitude is set as a threshold value for dividing the pixels in the block into two groups (s204), and the threshold value of the color signal having the maximum amplitude is determined. Then, the magnitude of the corresponding color signal is determined, and the pixels in the block are divided into two groups (s
205). As shown in FIG. 7B, this means that pixels distributed in the color space are divided into two groups according to threshold values.

In this example, since four types of colors are set, it is determined whether division of four groups corresponding to each color type has been completed (s206). If not, for each of the two groups divided in step s205, steps s203 to s
Grouping by 205 is performed to divide into four groups. That is, as shown in FIG. 7C, the color space is divided into four using three threshold values.

As described above, the group division processing for dividing into four types has been completed. Next, the processing proceeds to the code generation processing for each pixel signal. First, a selection signal S (a 2-bit signal representing 0, 1, 2, 3) for indicating which of the four groups belongs to each pixel mn is assigned (s207).

Then, for each group, the number of pixels having the same color among the pixels belonging to the group is measured, and the index of the maximum number is set as the representative color of the group, that is, the block index signal XB (s20).
8). As a result, as shown in FIG. 7 (d), the four types of colors used in the block are assigned to the block index signals XB0 to XB.
3 and an index selected by each pixel is represented by a selection signal S.
Can be represented by Finally, the selection signal S of each pixel and the block index signal XB are stored as compressed data in the compressed data storage device 13 (s209).

FIG. 8 shows a flowchart of an encoding process according to still another embodiment. In this example, the number of color types is variably set according to the image signal in the block. In the figure, steps s303 to s306 show a group division procedure, which determines whether or not division is possible using a preset amplitude range determination value, and sets the number of color types according to the number of divisions.

As an initial setting, parameters that determine the processing procedure such as the block size are set (s301), the pallet table 100 is read, and the image signal Xmn of the pixel in the block is converted into a three-color signal (RGB signal) (s3).
02).

Next, the maximum value and the minimum value are measured for each of the RGB color signals, the respective amplitude values are obtained, and the color type (R, G or B) having the maximum amplitude is obtained (s30).
3). Here, it is compared with the amplitude range determination value (s304),
If the maximum amplitude is small, it is expressed by one type of color signal, and the flow shifts to step s307. If this is the first processing result for the block, the color type of the block becomes 1 and the group division processing ends.

On the other hand, if the maximum amplitude is larger than the determination value, a threshold for dividing the pixels in the block into two groups is set (s305). The threshold value is an intermediate value in the amplitude range of the color signal having the maximum amplitude. Using the threshold value, the magnitude of the corresponding color signal is determined, and the pixels in the block are divided into two groups (s306).

Next, it is determined whether or not the group division has been completed (s307). If the possibility of the group division remains (N), the process returns to step s303, and the RGB division is performed again for each of the two divided groups. The maximum and minimum values are measured for each color signal, and the color type (R, G or B) having the maximum amplitude is determined for each group (s303). The maximum amplitude of each group is compared with the amplitude range determination value (s304), and the group having a small amplitude does not perform further division. On the other hand, the group with the largest maximum amplitude is again divided into two groups, and the pixels in the group are grouped (s305, s306).

As a result, when the second division processing for the block is completed, the color type becomes one of 2, 3, or 4. For the group divided in the second time, the process returns to step s303 and repeats the above processing. When the maximum amplitude of all the groups becomes smaller than the determination value, the division processing ends, and the number of color types of the block becomes This is the number of divided groups. To indicate the number of groups, the color type signal C
Set.

Here, the amplitude range determination value can be arbitrarily set according to a desired color type, and the upper limit can be four types. When the upper limit is four types, it is preferable to end the process by two division processes.

Next, in order to identify the group to which each pixel in the block belongs, a 2-bit selection signal S indicating the belonging group is assigned to each pixel (i, j) (s308). Then, the number of pixels having the same color among the pixels belonging to the group is measured, and the block index signal XB is set using the index of the color having the maximum number as the representative color of the group (s309). Finally, the color type signal C, the selection signal S, and the block index signal XB are stored in the storage device 13 as compressed data of the image signal (s31).
0).

As described above, the color used for each block group is represented by the block index signal XB, and the group to which each pixel belongs is represented by the selection signal S, so that the color signal of each pixel can be compressed.

Further, it is determined whether or not an adjacent portion having a temporal processing order or a two-dimensional arrangement or an adjacent block which has been converted into compressed data has the same color signal K set therein. Is set, the compression ratio can be further increased by substituting a signal representing the same as the adjacent part.

The encoding processing of the color image signal according to the present invention has been described through a plurality of embodiments. According to this, the number of color types used for each block including a plurality of pixels is fixed or variably set, and the palette index X of the representative color that appears most frequently in the block or the group within the block is associated with the block index signal XB.
For example, 256 color types (8 bits) can be represented by a palette within the upper limit of four types (2 bits) in a block, and a color image signal can be significantly compressed.

Further, according to the present invention, the variation range of the compression ratio can be set by the editing device 15 as a parameter. For example, when a palette table is composed of 256 types of color signals, and a block is 4 × 4 = 16 pixels and a 2-bit color type signal is added using four types of block index signals, the compressed data per block is The maximum is 66 bits, and the minimum is 10 bits. Since the data amount of the input image signal is 16 pixels × 8 bits = 128 bits,
The compression ratio can be set to about 1/2 to 1/12.

As described above, since the maximum value of the capacity of the compressed data to be generated can be calculated, the maximum value of the memory capacity for storing the compressed data, the time required for data transmission, the time required for the decoding process, and the like is obtained. , The configuration and performance of the device can be designed.

Further, according to the present invention, parameters can be adjusted so as to approach the target compression ratio. FIG.
FIG. 9 shows a flowchart of an encoding process in which a target compression ratio is set. In the figure, the same step number s as in FIG. 6 represents the same processing.

First, initial setting of parameters for code enable processing (s201), target compression ratio and allowable range, types of parameters used for adjustment, initial values of the adjustment parameters, and the like are set (s210). . Next, the encoding process s
Steps 202 to s208 are executed, and the compression ratio of the generated compressed data is compared with the target compression ratio (s211). Then, it is determined whether the target compression ratio falls within the allowable range (s21).
2).

As a result, if it is within the allowable range, the processing is terminated and compressed data is output (s209).
Use the set parameters for decryption. On the other hand, if the difference is not within the allowable range, the parameter is changed so that the difference becomes smaller, and the encoding process is executed again.

It should be noted that parameters can be set so that the average compression ratio of a plurality of image signals, instead of a single image signal, falls within the range. For example, in the case of a temporally continuous image such as a moving image, encoding, accumulation (or transmission),
Perform decryption processing.

By the way, according to the encoding processing of the above embodiment, since the type of color signal used for each block is limited, a pallet index (color) which is not referred to throughout the entire screen is generated. Therefore, the size of the pallet table can be reduced by deleting the contents that are not referred to.

The pallet table reconstructing section 110 uses two counters and two memory areas of the pallet table storage device 100, reconstructs the pallet table stored in the first memory area, and then reconstructs the second memory. Store in area. FIG. 10 shows a processing procedure of pallet table reconstruction. In this example, the third work area is used as a temporary work area.
Prepare a memory area of

First, in order to sequentially access the first memory area storing the palette table from the top,
A counter 1 pointing to a table index signal Pk;
The start address is set to the counter 2 as an initial value (s
401). The maximum value of counter 1 is pallet table 1
00 indicates the last address.

Next, the index signal Pk of the counter 1
Is compared with the block index signal XB of the compressed data (s402), and it is determined whether or not they match (s4).
03). If they match, it is determined that the index signal Pk is used for the compressed data, and the content (color signal) of the index signal Pk indicated by the counter 1 is read from the first memory area (palette table) (s40).
4) Write the read color signal to the second memory area indicated by the counter 2 (s405). And counter 1
And the value of the counter 2 are stored in the third memory area as the index signal before and after the change (s406).

Next, it is determined whether or not the counter 1 has reached the final value (s407). If not, the counter 1 and the counter 2 are counted up (s40).
8) Return to step s402. If the final value has been reached, the block index signal XB of the compressed data corresponding to the counter 1 is converted to the counter 2 using the pre-change and post-change index signals read from the third memory area.
(S409).

According to the above procedure, a palette table reconstructed by removing colors not used in the compressed data,
It can be generated simultaneously with the compressed data referring to the pallet table. The pallet table reconstruction unit 1
10 may be configured as one function of the editing device 15.

Next, the decoding process of the compressed data by the decoding device 14 will be described. FIG. 11 is a flowchart illustrating the decryption processing procedure. In this example, the same color signal K is not used, and the decoding process is performed on a single block.

First, initialization such as the number of blocks and preparation of a pallet table is performed, and the first block is designated (s
501). Next, the compressed data of the block is input, and the number of block index signals XB is determined from the color type signal C (s502).

Next, the block index signal XB and the selection signal S of the block are read (s503). And
For each pixel, the block index signal XB is selected from the selection signal S, the value of the signal XB is converted into a read address of the palette table, and the table contents are read to output the color signals R, G, B (s504). It is determined whether the block is a block (s505), and if there is a next block, the process returns to step s502 (s506), and the above process is repeated.

When the same color signal K is used, it is determined from the value of the signal K whether the block index signal XB of the adjacent block can be used. When the color signal can be used, the color signal of each pixel of the block is reproduced using the block index signal XB of the adjacent block.

The compressed data according to the present embodiment includes the same color signal K, the color type signal C, the block index signal XB, and the selection signal S. Here, it is noted that the number of bytes of the compressed data per block can be determined only from the same color signal K and color type signal C at the head. When the same color signal K is not used, the number of bytes of the compressed data can be determined only from the color type signal C. Therefore, when the compressed data of one screen is stored in the memory, by sequentially reading out the color type signals C for each block, the compressed data of an arbitrary block of the image can be searched at high speed.

In order to further shorten the search time, for example, for a block for each of a plurality of lines, a line head memory address is searched in advance and compiled in a table. Processing may be performed. In searching for the line head memory address, the color type signal can be sequentially detected simultaneously with the input of the compressed data. In this case, the processing time for the search can be further reduced. Thus, while the compressed data of this embodiment is a variable length code,
It has the feature that partial images can be searched at high speed from compressed data.

FIG. 12 shows an example of the configuration of a decoding device. The decoding device 14 shown in FIG. 3A temporarily stores the input compressed data in the buffer 21, reads out the compressed data in block units, sets the read data in the register 22, and the selection circuit 23 outputs a selection signal for each pixel. S, one of the block index signals is selected, the palette table 100 is searched using the block index signal, and the corresponding color signal (R
GB).

The decoding device 14 shown in FIG. 8B reads out the selection signal S for each pixel in block units from the buffer 21 temporarily storing the compressed data, and
The color signal (RGB) of the block is searched from the palette table 100 using the block index signal XB and set in the register 22-2. Selection signal S
To select and output the corresponding color signal.

Since the decoding process is performed on the compressed data obtained by the encoding process of the present invention, the number of data to be read from the compressed data storage device 13 is small, and the data bus connected to the storage device 13 for the decoding process is small. The above data transfer load can be reduced.

Further, in the decoding method shown in FIG. 12B, the color signals used in units of blocks are read from the pallet table 100, set in the register 22, and used in the block by using the selection signal S for the pixels in the block. Since the color signal to be used is selected, it is not necessary to search the entire pallet table, so that the number of accesses to the pallet table 100 is reduced, and the number of address lines operated in pixel units can be reduced.

According to the present embodiment, since the number of times of memory access in the decoding device can be reduced, the memory for storing the compressed data and the memory for storing the pallet table are made identical so that the memory access timing does not overlap. And the number of circuit elements can be reduced. In addition, program memory and work memory area, etc.
It can be configured the same as the above memory means.

FIG. 13 shows a specific function of the palette color expansion circuit. The illustrated configuration is an application example of the decoding method in FIG. The input compressed data is temporarily stored in the buffer 21 in order to match the signal input / output speed.

For example, if the compressed data format is 4 × 4 pixels,
When the color signals used in the block are limited to four colors, the compression ratio is fixed, so that the memory address of the compressed data is regularly determined based on the order of the pixels. Accordingly, the block index signal XB and the selection signal S are read out from the buffer 21 for each block from the buffer 21 using the outputs of the pixel counter 24 and the line counter 25, which count up in synchronization with the pixel clock, and stored according to the content of the compressed data. The value is set in the register 22 while selecting the destination. Then, the selection signal Smn corresponding to the output pixel Xmn is selected using the pixel counter 24 and the line counter 25, and one of the block index signals XB is selected using the selection circuit 23. Further, the pallet table 100 arranged at the subsequent stage is searched using the selected block index signal XB, and a color signal (RGB) of the corresponding index X is output.

When the image output destination is a display, RG
To output a signal for each color plane of B, the conversion result of the palette table 100 is selected and output based on the color plane signal. The color plane signal can be calculated from the number of screens measured by the pixel counter 24 and the line counter 25. On the other hand, if the image output destination is a printer, YMC converted from RGB signals
Printing is performed using K (yellow, magenta, cyan, black). Therefore, by simultaneously outputting RGB signals, color conversion from RGB to YMCK is performed.

FIG. 14 shows a compressed data decompression circuit for a natural image according to Reference Example 2. The compressed data of the cited example 2 does not use a palette color, but instead includes a color signal approximating an image signal in a block and a selection signal of each pixel.

The decompression circuit includes a buffer 21 'for temporarily storing input compressed data, a register 22' for setting a selection signal and an approximate color signal for each block, and a register 2 '.
There is a selection circuit 23 'for selecting one of the approximate color signals set to 2' for each pixel. The output of the pixel counter 24, which counts up in synchronization with the pixel clock, and the output of the line counter 25 are used as address signals,
The approximate color signal and the selection signal S are read from 1 for each block, and set in the register 22 while selecting a storage destination according to the content of the compressed data. Then, the selection signal Smn corresponding to the output pixel mn is selected using the pixel counter 24 and the line counter 25, and one of the approximate color signals is selected using the selection circuit 23.

Further, the gamma table 200 is searched using the selected approximate color signal, and the corresponding color signal is output. The gamma table is used as signal conversion means for performing fine color adjustment. Therefore, each of the pallet table and the gamma table is, for example, an 8-bit signal (2
56 types) and 8 bits (24 bits in total) of three colors of RGB are output, so that both can have the same configuration.

It should be noted that the basic configuration of the decompression circuit is almost the same in FIG. 13 for the palette color and in FIG. 14 for the natural image. The difference between the two is that the former is a pallet table 100, which is an essential element, and the latter is a gamma table 200.

As described above, when the basic part of the decompression circuit is designed to be able to be shared with the compressed data of the palette color or the natural image, the decompression device applicable to both the compressed data and the image display device including the same can be realized. Will be possible. Furthermore, if the compression method is switched for each block depending on whether the target image is a natural image or a palette image, it can be applied to a case where a natural image is mixed in a part of the pallet type screen. The rate can be greatly improved.

Next, an example in which the present invention is applied to an image display device will be described. As an image display device, a printer, a CRT, a liquid crystal display, a plasma display, or the like can be used. In each case, an image signal is formed by performing line-by-line scanning. In order to supply image data to these image display devices, normally, image data for one screen is stored in a bitmap memory, a memory address is generated according to a scan order, and the corresponding memory contents are read and transferred. do it.

By the way, image data becomes enormous in volume as high definition and high gradation images are obtained. On the other hand, writing / reading of a memory for storing image data is restricted by an access time determined by characteristics of a memory element. In addition, if a high-speed memory or a wide data bus is used for high-speed data input / output, the cost of the device increases and noise is generated.

In the present invention, in order to solve the above problem, the image data of the image display device is stored in the memory by the compressed data, and the memory capacity and the data transfer rate are reduced. As described above, the compressed data of this embodiment is m × n
, The same block can be read and decoded over n lines, and image data can be output without delay. Hereinafter, the configuration and operation of the image display device according to one embodiment will be described in detail.

FIG. 15 shows the configuration of the control unit of the image display device. In the case of an image display device, since scanning is performed in units of lines, it is necessary to convert the configuration of compressed data in blocks of m × n pixels into a configuration in units of lines.

The display control device 500 includes a register 501 for reading and storing at least one block of one line of compressed data, a memory address generation circuit 502 for reading compressed data from the compressed data storage device 13, and decoding of compressed data. And a synchronizing circuit 503 for synchronizing reading and decoding processing with scanning of a display device (not shown).

FIG. 16 shows the relationship between scan lines, blocks, and pixels. The illustrated example is for simplicity of explanation.
One block is 4 × 4 pixels, and blocks arranged in the line direction of one screen are the first to third blocks. Pixel P (i, j)
Is located at the left end of the first line of the first block at P (1,
1), the right end of the fourth line is P (4,4). Note that the relationship between the scan line number, the memory read address, and the pixel position of the decoded image data is uniquely determined.

FIG. 17 is a timing chart showing the operation of each part of the display control device. FIG. 9A shows the operation for the first line. (A) Memory address generation circuit 502
, The address is calculated from the first block at the upper left of the screen, (b) compressed data of the first to third blocks is read, and (c) writing to the register 501 is performed. Here, the register 501 is composed of a replacement buffer,
The data of the block is written in the first register, and the data of the second block is written in the second register, alternately with the block, and the process proceeds without interruption.

Next, (d) the decoding circuit 507 generates the corresponding block index signal based on the selection signal S (1,1) which is the compressed data of the pixel P (1,1) on the first line of the first register. XB is selected, and the color signal of the pixel is searched and decoded from the palette table 100, and (e) the D / A converter 508
And outputs a color signal to the display device. Similarly, color signals are sequentially output to the pixels P (2,1) and P (3,1) on the first line of the first block.

When the first block is completed, pixels P (5,1) to P (8,1) of the second block and pixels P (9,1) to P (12,1) of the third block , The pixels of the first line are sequentially and continuously output, and the drawing of the first line on the display device is performed.

When the first line is completed, the first to third blocks are read again, and the same processing is repeated for the second line. Further, when the processing up to the fourth line is completed, the same processing is repeated for the first to third blocks arranged in the next line direction on the screen.

As described above, in synchronization with the scan of the display device, only the pixels of the same line in the block are decoded from the compressed data, and the obtained one-line image signal is immediately output and displayed. The device configuration is simplified, for example, a bitmap memory is not required. In the conventional JPEG encoding method, in order to collectively decode compressed data of a block, a memory for temporarily storing image data obtained in a block unit and a signal of a necessary pixel are selected from the memory. I needed a way to do it.

Next, the conversion method of the compressed data format will be described. As described in the above encoding method, the number of signal types used in a block is fixed in advance, or is variable by a threshold value of the frequency of appearance of a color signal or the like under a limit smaller than the number of pixels in the block. can do. A variable compression data format can increase the compression ratio compared to a fixed compression data format,
The decoding process takes time. Thus, the variable format is converted to a fixed format before decoding, so that the decompression process in the image display device can be speeded up.

FIG. 18 shows the structure of compressed data for four blocks in the first format and the second format. FIG. 7A shows a variable-length data structure according to the first format, which has a color type signal C indicating the number of color types of each block at the head, and stores compressed data having different data lengths for each block. I have. Incidentally, the block index indicating the color type of the A block is two of XB _A0 and XB _A1 , and the B block is X
The B _B0 and C blocks are each one of XC _C0 , and the D blocks are four of XD _D0 , XD _D1 , XD _D2 and XD _D3 .
Since there is only one color type for the B and C blocks, there is no selection signal.

FIG. 13B shows a fixed-length data structure in the second format, which does not have the color type signal C, and each block has four block index signals according to the longest data in the first format. And selection signals corresponding to the pixels in the block. Here, the block index signal XB that does not exist in the first format may be a dummy because it is not adopted in the second format.

FIG. 19 shows functional blocks of the data format converter. The compressed data read from the first memory (for example, the memory 13) in units of a predetermined number of blocks (here, four blocks) by the memory address generation device 502 are respectively stored in the selection signal register 22-1 and the block index register 22. -2, color type signal register 2
Stored in 2-3. The conversion data generation means 601
It detects the data length of each block from the color type signals C _A -C _D blocks, and outputs to the memory address generator 502. Also, in the second memory (for example, the buffer 21),
The block index signal and the selection signal for each block are stored. At this time, a dummy is represented in the block index signal area which is not set to a variable length, and the selection signal is represented in a maximum length format regardless of the color type of the block.

FIG. 20 shows an example of variable-length data and an explanatory diagram of decoding after fixed-length conversion. FIG. 3A shows the relationship between blocks A to D and pixels on the screen, and FIG. 3B shows the compressed data structure of each block. Since the selection signal of the block A has two kinds of colors, the selection signal is binary with one bit. Since there are four types of blocks D, the selection signal is a four-bit alternative with two bits. Therefore, the conversion data generating means 60
All of the compressed data converted by 1 is stored in the second memory in the form of a block D.

As a result, the decompression device of FIG.
As shown in (c), a 2-bit selection signal S is provided for each pixel.
_{Based on ij} , block index signals XB0 to XB3 are alternately selected, and XB corresponds to the corresponding pallet table 1
From the index P _{k of} 00, the corresponding color signal (RGB)
Is output.

According to this, since the compression ratio is fixed regardless of the pattern, the position of the block (or pixel) in the image can be uniquely associated with the memory address where the compressed data is stored, and the readout can be performed. Processing can be accelerated.

Reading of compressed data from the first memory is performed every n lines. Then, in synchronization with the timing of consuming one line of data in the second memory, the data format is converted from the first memory to the second memory and the compressed data is transferred. As a result, there is a margin in the data input / output timing of the first memory, so that it becomes possible to write new compressed data or input / output other data.

When the bitmap memory system is adopted, the decoding process is executed during the data transfer from the first memory to the second memory, so that n
Line bitmap data can also be stored.

As described above, when the encoding / decoding processing of the present invention is applied to an image display device, in addition to the above-described partial image retrieval processing, various image displays can be speeded up and circuits can be simplified. Can be realized. Hereinafter, examples of image rotation, enlargement / reduction, and synthesis of a plurality of screens will be described.

When executing processing such as rotation and enlargement / reduction of an image signal, the image signal is stored in the memory as compressed data, and the compressed data is read out in block units according to a scan order based on the rotation / enlargement / reduction. Then, signal processing such as rotation and enlargement / reduction can be performed on the image signal of the block. Here, if attention is paid to the fact that the compressed data depending on the position in the block is only the selection signal S and the other compressed data is commonly used in the block, the rotation, enlargement / reduction, etc. Can be executed for the selection signal S.

That is, regardless of the image signal reproduced by the decoding process, the rotation of the image for the compressed data,
Enlargement / reduction can be realized as follows.

(A) The position signal of the next image signal to be output is calculated from the selection signal S based on the scan order and conditions such as rotation and enlargement / reduction, and the memory address of the compressed data of the corresponding block is calculated.

(B) The compressed data of the corresponding block is read from the memory and stored in a temporary buffer.

(C) The image signal X at the pixel position to be output is
Decoding is performed from the compressed data stored in the temporary buffer.

(D) Perform high-quality processing (eg, color conversion, gamma conversion, level conversion) based on the characteristics of the printer, and output a signal.

(E) The above procedure is executed until printing of one screen is completed.

Further, as an example of synthesizing a plurality of images, for example, a document image in which characters, figures, natural images, and the like are created and combined in different image planes, and roads, tracks, topography, and the like are created and combined in different image planes. There is a map image to do.

A plurality of image signals to be synthesized are prepared, for example, by reading from a CD-ROM or by creating drawing data using a processor. The order of superimposition of a plurality of image planes for the synthesis processing includes, for example, a method of assigning priorities to the image planes themselves, and a Z-buffer method of designating an image plane having a priority for each pixel.

The process of synthesizing a plurality of images can be realized by the following procedure using compressed data according to the encoding method of the present invention.

(A) Prepare a pallet table commonly used for all image planes to be combined. The color signal of the background is assumed to have the same value over the entire screen, for example, white.

(B) The horizontal / vertical synchronization signals of the display device are input, and the memory addresses of the compressed data are generated in accordance with the scan order.

(C) The compressed data (block index signal, selection signal, etc.) of the block to which the display pixel belongs is read from the memory and temporarily stored in a register. The register can be configured to accumulate compressed data for each image plane to be combined, and it is assumed that the priority of each image plane is assigned to each register.

(D) An image plane having significant color signals excluding the transmission color and the most background color and having the highest priority is selected. However, if there is no image plane having a significant color signal, the most background is selected.

(E) Using the selection signal S of the corresponding pixel on the selected image plane, the corresponding block index signal XB
Is selected, the palette table is read using the result, and a color signal is output. However, when the highest background is selected, the highest background color is output as a color signal.

(F) The procedure from step (b) is repeated until the scanning of the screen is completed.

As described above, by performing the synthesizing process based on the priority using the compressed data based on the pixel selection signal and the block index signal, the amount of data to be processed is reduced. Can be simplified, the number of wirings can be reduced, and the number of times of reading the image signal from the memory can be reduced, so that the speed can be increased.

It is to be noted that an image signal (for example, a natural image of 8 bits for each color) representing each pixel as a three-color signal can be synthesized with the image plane and output. For example, when displaying a natural image on a part of a display screen, an image plane for writing an image signal of the natural image is prepared, and the highest priority is assigned to the plane. A color signal is written to the pixels in the display area, and a transmissive color is written to the other areas, thereby performing a combining process with another image plane. The image signal on the natural image plane may be output without performing the decoding process.

Further, when a partial image is cut out from a large image plane such as a map image and displayed while being scrolled, the image signal of the image plane is composed of the above-described compressed data, so that the data transfer load can be reduced. The effect is great when applied to various image display processes.

[0132]

According to the present invention, a high compression rate can be realized by allowing a signal change that does not affect the image quality. When the image signal is in units of 8 bits, the compressed data can be grouped in units of bytes (8 bits), so that data input / output and signal processing can be simplified. Image quality and compression ratio can be easily adjusted using parameters such as color type signals. Further, since the decoding process is simple, a high-speed decoding process synchronized with the operation of the display device is possible. Furthermore, since the basic configuration of the decompression device can be made substantially the same for the compressed data of the pallet image and the natural image by block division, a sharable image processing device can be realized.

[Brief description of the drawings]

FIG. 1 is a flowchart showing a color image signal encoding method (first embodiment) of the present invention.

FIG. 2 is a configuration diagram of a color image processing apparatus according to an embodiment of the present invention.

FIG. 3 is an explanatory diagram showing an image signal and a compressed signal in block units and a data structure of a palette table.

FIG. 4 is an explanatory diagram showing a data structure of compressed data according to one embodiment.

FIG. 5 is an explanatory diagram showing a memory arrangement of compressed data according to one embodiment.

FIG. 6 is a flowchart illustrating an encoding processing procedure according to a second embodiment.

FIG. 7 is an explanatory diagram showing a method of determining a color signal used in a block.

FIG. 8 is a flowchart illustrating an encoding processing procedure according to a third embodiment.

FIG. 9 is a flowchart showing a procedure for realizing a target compression ratio in the encoding procedure according to the second embodiment.

FIG. 10 is a flowchart illustrating an example of a processing procedure for reconstructing a pallet table.

FIG. 11 is a flowchart showing a decoding processing procedure according to an embodiment of the present invention.

FIG. 12 is a configuration diagram of a decoding device according to an embodiment.

FIG. 13 is a configuration diagram of a decompression circuit using a palette color image.

FIG. 14 is a configuration diagram of a decompression circuit based on a natural image.

FIG. 15 is a configuration diagram of an image display device according to one embodiment.

FIG. 16 is an explanatory diagram showing the relationship between scan lines, blocks, and pixels.

FIG. 17 is a timing chart illustrating the operation of the display control device.

FIG. 18 is an explanatory diagram showing a memory arrangement of compressed data in a first format (variable length) and a second format (fixed length).

FIG. 19 is a functional block diagram showing a compressed data format conversion device.

FIG. 20 is an explanatory diagram showing compressed data of variable length and decoding after conversion to fixed length.

[Explanation of symbols]

10 image signal input device, 11 image signal storage device, 1
2 ... encoding device, 13 ... compressed data storage device, 14 ... compressed data decoding device (decompression device), 15 ... editing device, 1
6 output device, 21 buffer, 22 register, 11
DESCRIPTION OF SYMBOLS 1 ... Selection circuit, 24 ... Pixel counter, 25 ... Line counter, 100 ... Palette table, 101 ... Palette table reconstruction part, 200 ... Gamma table, 500 ... Display control device, 501 ... Register, 502 ... Memory address generation circuit, 503: Synchronous circuit, 504: Scan address generating circuit, 507: Decoding circuit, 508: D / A converter, 601: Conversion data generating means

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁶ Identification code FI H04N 1/46 H04N 1/40 D 11/04 1/46 Z

Claims (15)

[Claims] 1. A method of encoding a color image signal in which a color signal of each pixel of a color image is represented by a palette color registered in a palette table, wherein an image for one screen is divided into a predetermined number (m × n) of pixels. For each block divided so as to include, the number of color types used in accordance with the appearance of the palette color of each pixel in the block is limited to a predetermined number smaller than the predetermined number, and for each color type used, A block index signal for referring to a pallet table, and a selection signal for associating one of the block index signals with each pixel when the plurality of block index signals are set, wherein the color image signal is based on the selection signal and the block index signal. A color image signal encoding method characterized by being represented by compressed data. 2. The method according to claim 1, wherein the predetermined number of types of colors to be used is a fixed value set in advance in descending order of appearance frequency of palette colors in the block, or a lower limit value set in advance in the appearance frequency. A method for encoding a color image signal, wherein the variable value is equal to or less than the fixed number. 3. The color signal of each pixel according to claim 1, wherein the predetermined number of color types to be used is obtained by obtaining a maximum amplitude of a color signal component of a pixel in a block and using an intermediate value as a threshold value. After that, the maximum amplitude of the component of the color signal of the pixel in the group is determined, the intermediate value thereof is set as a threshold, and the color signal of each pixel is again subjected to the grouping process a predetermined number of times, or When the maximum amplitude is equal to or greater than a set amplitude value, the grouping is advanced, and when the maximum amplitude is less than the set amplitude value, the grouping is stopped, and the number of variable groups obtained when the division ends is determined. Characteristic color image signal encoding method. 4. The compressed data according to claim 1, 2 or 3, wherein the compressed data is provided with a color type signal indicating the number of color types used in the block, and is configured for each block or each of a plurality of blocks. Encoding method of a color image signal to be described. 5. The compressed data according to claim 4, wherein a same color signal indicating whether or not an adjacent block and the block index signal are the same is added to the compressed data, and the block index signal is added only to non-identical blocks. A method for encoding a color image signal. 6. A method of decoding a color image signal by decompressing compressed data of a color image signal by a palette color method and representing a color signal of each pixel by a palette color registered in a palette table. Is divided into blocks each containing a fixed number (m × n) of pixels, a block index signal for referring to the palette table for each of a predetermined number of color types used in the block, and a block index signal for each pixel. When the compressed data is configured from the selection signal (S) corresponding to the signal, the compressed data is read out in block units, and a block index signal corresponding to each pixel is selected from the selection signal in pixel order. Searching the palette table using a block index signal and decoding to the corresponding palette color Decoding method for a color image signal, characterized. 7. The compressed data according to claim 6, wherein the compressed data is sequentially read in block units from a first block to a last block arranged in a line direction of the screen.
After repeating the process of continuously decoding the compressed data of the same line of each block up to the last block for n lines, the decoding process from the first block arranged in the line direction at the next stage of the screen to the last block is performed. A method for decoding a color image signal, characterized by shifting. 8. The decoding device according to claim 7, wherein the second block is read while the read compressed data of the i-line (i ≦ n) of the first block is decoded, and the decoding of the i-line of the first block is completed. A decoding process for starting decoding of the i-line of the second block, and continuously outputting the i-line color image signal until the last block. 9. An image signal storage device for storing a color image signal created or input by a palette color system, a compressed data storage device for storing compressed data of the color image signal, a predetermined number of palette colors An image processing apparatus comprising: a pallet table for adding an index to and storing the data; an encoding unit for encoding the color image signal to create the compressed data; and an editing unit for controlling each unit. Is a fixed number of images for one screen (mxn)
For each divided block including the pixel, the palette table is referred to for each color type limited to a number smaller than the certain number used in the block according to the appearance of the color type by the color image signal of each pixel. Color image processing by setting a block index signal for performing the operation and a selection signal that associates the color type of each pixel with the block index signal, and creating the compressed data based on the selection signal and the block index signal. apparatus. 10. The block index signal according to claim 9, wherein the block index signal read out from the compressed data storage unit for each block and the storage means for storing the selection signal, and one of the block index signals stored using the stored selection signal. A color image processing device comprising: a decompression device including a selection unit for selecting a color signal and a decoding unit for extracting a corresponding color signal from the palette table using the selected block index signal. 11. A block index signal representing a plurality of approximate colors within a block and a selection of each pixel for each block obtained by dividing a natural image including a fixed number (m × n) of pixels. When compressed data of a natural image is created from a signal, a gamma table for searching for the approximate color is further provided, and the decompression device can be shared according to a palette image or a natural image. A color image processing device characterized by the above-mentioned. 12. The color image processing apparatus according to claim 9, further comprising a palette table reconstructing unit that deletes the contents of the palette table that is not referred to by the encoding unit or the decoding unit. . 13. The block index signal according to claim 10, 11 or 12, wherein one of the block index signals is defined as a transparent color, and a signal indicating the context of a plurality of images and a compressed data of the plurality of images are located at the same position. A color image processing apparatus comprising: an image display device that receives a selection signal of a certain pixel and outputs a selection signal for displaying the pixel, and that determines an anteroposterior relationship and combines and displays a plurality of image signals. apparatus. 14. A compressed data storage device for storing compressed data of a color image signal according to a palette color system, a palette table for storing a predetermined number of palette colors with an index, and decompressing the compressed data so that each pixel is expanded. An image processing apparatus comprising: a decompression device that outputs a color image signal in a palette color; and a display device that scans and displays a decoded color image signal in a line direction of a screen. Is a fixed number (mxn)
For each divided block including pixels, a block index signal that refers to the palette table for each variable or fixed color type used in the block, and a pixel order that associates the color type of each pixel with the block index signal. The display device issues a read address of the compressed data according to a synchronization signal of a line scan and a fixed length or a variable length of the compressed data when the selection device is configured from a selection signal and stored in the compressed data storage device. And the first lined up in the line direction of the screen
From the block to the last block, the compressed data is sequentially read out in block units, and the compressed image of the same line in each block is decoded by repeating the processing by the decompression device from the block to the last block continuously until the last block. A color image processing apparatus for displaying a signal, repeating the processing from reading to display for n lines, and then repeating the same processing for each block arranged in the line direction at the next stage of the screen. 15. The color image processing apparatus according to claim 14, further comprising a data format conversion unit that converts fixed-length data when the compressed data has a variable length.