JPH08154261A - Image compressor - Google Patents

Abstract

PURPOSE: To obtain an image compressor for compressing an image into the resulting compressed image, which is easily expanded by an image expander. CONSTITUTION: A division section 121 acquiring image data of a luminance signal Y and color difference signals U, V divides the data into 4×4 blocks and an encode section 122 generates two colors of 1st color component (y1, u1, v1) and 2nd color component (y2, u2, v2) and a bm representing the distribution as to each block. A color extinct section 12 generates a CLUT from the color components. Block data B of most frequent incidence are detected as Rb and the data are sent from a transmission section 14. Then two kinds of color components are extincted to be C1, C2 based on the CLUT, a compression section 13 compares a processing object block Cb with a reference block Rb and a surrounding block to code the data into a code of the equal block. In the case of inequality, the data are coded based on code information and difference data of a most similar block. The coded data are sent from a transmission section 14.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image compression device, and more particularly, to an image compression device for displaying a color image by color reduction by CLUT.

[0002]

2. Description of the Related Art When outputting image data to an image display device, by lowering the data transfer rate of the system bus,
In order to reduce the load on the CPU (Central Processing Unit),
Color reduction processing may be performed. The color reduction process is, for example, 1
This is a process of displaying 256 or 16 colors of a full-color natural image of 2 ²⁴ = 16.77 million colors in which 8 bits of RGB (red, green, blue) or YUV (luminance and color difference) are assigned to pixels. In this color reduction processing, 256 colors or the like are selected from among 16770,000 combinations of colors, and a CLUT (Color Look Up Table) in which each color is indexed
Create This CLUT is used as pixel information.
It corresponds to the index of.

On the other hand, color moving images and still images are digitally signal processed and transmitted / received in video conferences, video telephones, and online karaoke. Since the data amount of such a color image is very large, it is important to perform simple and efficient compression when transmitting the data. As a conventional compression method of image data, for example, there is a discrete cosine transform (DCT), which is JPEG (Joint Committee of Photography Specialists).
It is used when compressing a color still image according to. The pixel data after the image compression / decompression processing by JPEG or the like has color value data, and the pixel value on the video memory and the color value correspond linearly. For this reason, J
When applying compression / decompression processing by PEG, etc., the decompression side performs decolorization processing on the decompressed pixel data and CL
I try to create a UT.

[0004]

As described above, when index color display and JPEG or the like are applied, the decompression side must perform color reduction processing to create a CLUT, which makes the software processing of the decompression side device heavy. However, there is a problem that the hardware configuration becomes complicated. Therefore,
It is an object of the present invention to provide an image compression device that performs image compression that can be easily decompressed by a device on the image decompression side.

[0005]

According to a first aspect of the present invention, image data acquisition means for acquiring image data in which each pixel data is composed of a luminance signal and a color difference signal, and the image data acquisition means. A block generation unit that generates block data composed of n × m pixels from image data, and a first color component that has a luminance signal and a color difference signal that represents the block from the block data generated by the block generation unit. The second color component, and the first color component and the second
Encoding means for generating a block component made up of an n × m bitmap showing the distribution of color components, and color reduction processing to an arbitrary number of colors for the first color component and the second color component generated by this encoding means. The image compression apparatus is provided with a CLUT creating means for creating a CLUT for performing the above, and a sending means for sending the CLUT created by the CLUT creating means and the block component created by the encoding means. To achieve. According to a second aspect of the present invention, in the image compression apparatus according to the first aspect, the first color component and the second color component of the block component generated by the encoding means are used.
CL created by the CLUT creating means for color components
A color reduction unit that performs color reduction processing by the UT to generate a block component composed of a color-reduced first color component, a second color component, and an n × m bitmap, and the transmission unit includes the CLUT.
The CLUT created by the creating means and the color-reduced block component created by the color-reducing means are transmitted. According to a third aspect of the present invention, in the image compression apparatus according to the first or second aspect, the block component generated by the encoding unit or the block component after color reduction generated by the color reduction unit is used as the block component. And a coding unit that selects a similar block whose block components are the same or is most similar and that codes from the code data indicating this similar block and the difference data with respect to the block component of the similar block, and the transmitting unit. , The CLUT created by the CLUT creating means and the coded data after being coded by the coding means. According to a fourth aspect of the invention, in the image compression apparatus according to the first to third aspects, the CLU is used.
The image data acquisition means acquires the image data of one frame for the still image, and includes a multi-frequency block detection means for detecting a block having a high frequency of the color-reduced block components after the color reduction processing by the CLUT created by the T creation means. The encoding means selects a block that matches or includes the block component of the block detected by the multi-frequency block detection means, and the most similar block, and the transmission means further includes the multi-frequency block detection means. Send the block component of the detected block.

[0006]

According to the image compression apparatus of the first aspect, each pixel data obtains image data composed of a luminance signal and a color difference signal, and generates block data of n × m. Then, a block composed of a first color component and a second color component having a luminance signal and a color difference signal, and an n × m bitmap showing a distribution of the first color component and the second color component, which represents the generated block. Compress the data by generating the components. In addition, a CLUT for performing color reduction processing on the first color component and the second color component to an arbitrary number of colors is created and transmitted together with the block component. The image compression apparatus according to claim 2, wherein the first color component and the second color component of the block component generated by the encoding means are subjected to color reduction processing by the CLUT created by the CLUT creating means, and the color-reduced A block component composed of one color component, a second color component, and an n × m bitmap is generated. Then, the CLUT and the block component after color reduction are transmitted. The image compression apparatus according to claim 3 selects a block component that has not been reduced in color or a block component that has been reduced in color and a similar block whose block component matches or is most similar to the block component, and indicates this similar block. The data is further compressed by encoding the code data and the difference data for the block component of the similar block. The image compression apparatus according to claim 4 detects, in a still image, a block in which the frequency of the color-reduced block component after the color-reduction processing by the CLUT is high, and includes a block component of this multi-frequency block, or a block of the most similar block. By making a selection, the compression rate is further increased.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An image compression communication system using an image compression apparatus according to an embodiment of the present invention will be described in detail below with reference to FIGS. FIG. 1 shows the configuration of an image compression communication system for transmitting and receiving image data. As shown in this figure, the image compression communication system
The transmission side shown in FIG. 1A compresses / encodes image data and transmits it, and the reception side shown in FIG. 1B receives and decodes / decompresses compressed data. On the transmission side of the image compression communication system, each pixel has a luminance signal Y,
A YUV acquisition unit 11 that acquires image data having color difference signals U and V, a block encoding unit 12, and a color reduction unit 15.
And a compressing unit 1 for determining and coding inter-frame and intra-frame coincidence of the reduced color block data B.
3 and the transmitter 14.

The block encoding unit 12 divides the image data acquired by the YUV acquisition unit 11 into blocks of a predetermined n × m size, and a block data of the blocks from the YUV data of each pixel forming the block. The encoding unit 122 for generating A is provided.
The block data A of each block generated by the block encoding unit 12 is supplied to the color reduction unit 15. The color reduction unit 15 creates a CLUT from the block data A for all blocks in one frame C
An LUT creation unit 151 and a color reduction processing unit 152 that creates color reduction block data B from the created CLUT and block data A are provided. At the time of creating the CLUT, the CLUT creating unit 151 detects the most frequent color-reduction block data in the color-reduction processing, supplies this to the compressing unit 13 as the reference block Rb, and supplies the created CLUT to the transmitting unit 14. It has become.

The compressing unit 13 subtracts the reduced color block data B for the current frame which is the object of the encoding process.
For two lines, a reference block buffer 132 for storing the reference block Rb supplied from the color reduction unit 15, and data encoding by encoding according to whether the components of the color reduction block data match or do not match. The encoding unit 133 is provided. Transmitter 14
Sends the reference block Rb and the CLUT, and also sequentially sends the encoded data.

On the other hand, the receiving side receives the data transmitted from the transmitting section 14, the receiving section 24, the decompressing section 23 for inversely converting the encoded block data into the color-reduced block data, and the color-reduced block data for each pixel. A block decoding unit 22 for inverse conversion into color-reduced data and YUV
The output unit 21 outputs data to a display device or the like. The decompressing unit 23 separates the data received by the receiving unit 24 into a reference block Rb, a CLUT, and encoded data to be sequentially received, and a decoding unit to decode the encoded data into original block data. A block 231, a line buffer 232 in which the decoded block data for the line currently being decoded is stored for two lines, and a reference block Rb.
Is stored in the reference block buffer 233.

Next, the basic concept of image compression in the image compression communication system of this embodiment will be described. That is, in image compression communication systems, the human eye is sensitive to changes in brightness but insensitive to changes in color in image compression, and when the pixels that make up an image are appropriately blocked, many natural phenomena occur. The image compression is performed based on the premise that there are only two colors in one block that is substantially established in the image.

Next, the operation of transmitting and receiving image data by the embodiment configured as described above will be described. On the transmission side of the image compression communication system, the YUV acquisition unit 11 transmits image data of each frame represented by a luminance signal Y and color difference signals U and V for each pixel to, for example, a CCD (Charge Coupled).
Device)) etc. When color information consisting of R (red), G (green), and B (blue) is input to the YUV acquisition unit 11 as color information, YUV data is obtained from this color information according to the following formula 1. YUV is obtained by calculating. Each color information of RGB is generally represented by 8 bits, and 2 ⁸ × 2 ⁸ × 2 ⁸ = 16,777,216 colors are represented. Y, U, and
Each V data is also represented by 8 bits. In this way,
The Y, U, and V data acquired by the YUV acquisition unit 11 are sequentially supplied to the block encoding unit 12 in frame units.

[0013]

## EQU00001 ## Y = 0.299R + 0.587G + 0.114B U = 0.564 (BY) V = 0.713 (RY)

When the image data is supplied to the block encoding unit 12, the dividing unit 121 divides the image data into n × m such as 4 × 2, 2 × 4, 8 × 4, 8 × 8 for each frame. Divide into any size. In this embodiment, as shown in FIG. 2, each block is divided into 4 pixels × 4 pixels.
The resolution per original frame is arbitrary as long as it is an integral multiple of the division size, but in the present embodiment, 640 pixels × 480 pixels = 307,2 per frame.
It has a resolution of 00 pixels. Therefore, one frame of the original image is divided into 160 blocks × 12 by the dividing unit 12.
It is divided into 0 block = 19,200 blocks.

Then, the encoding unit 122, as shown in FIG. 3, extracts the block luminance component y from the YUV data for 16 pixels included in each divided block.
Block data A composed of 1, y2, block color difference components u1, u2, v1, v2, and bm (bitmap) is generated. Each component of this block data A is y
1, y2, u1, u2, v1, v2 are composed of 8 bits, and bm is composed of 16 bits. Thereby, the dividing unit 121
Data that was composed of 8 bits × 3 components (YUV) × 16 pixels = 384 bits per block when divided by 8 bits × 6 types (y1, y2, u1, u2, v1) by encoding. , V2) +16 bits (bm) = 64 bits. That is, the encoding unit 122 compresses the data to 64/384 = 1/6.

The block components y, u, v and bm are y
A first color component consisting of 1, u1, v1 and y2, u2, v
The second color component consisting of 2 is classified into bm, and the color reduction processing described later is performed on the two colors of the first color component and the second color component.
Here, bm is a map for indicating whether the 16 pixels forming each block take the first color component or the second color component, and the pixel taking the first color component is represented by "1". , A pixel that takes the second color component is represented by “0”.

Here, the calculation of each block component will be described. First, bm for distinguishing the first color component from the second color component
The calculation of will be described. As described above, bm is obtained from the luminance components Y1 to Y16 because the human eye is sensitive to luminance changes but insensitive to color changes.
That is, the threshold H is calculated from the pixel luminance components Y1 to Y16 in the block, and the luminance component Y larger than the threshold H is calculated.
A pixel having h is a first color component, and a pixel having a luminance component Yl smaller than the threshold value H is a second color component. The selection of the threshold value H is arbitrary and can be determined by various methods. In this embodiment, the largest value Ymax of the luminance components Y and the smallest value Ymin are excluded.
The average value of the individual luminance components Y is adopted as the threshold value H.

The average value of pixels having a luminance component Yh larger than the threshold value H is y1, and the average value of pixels having a luminance component Yl smaller than the threshold value H is y2. As for the block color difference component, the average value of the color difference components U (Yh) and V (Yh) of the pixel that becomes the first color component is u1, v1, and the color difference component U (Yl) of the pixel that becomes the second color component, The average of V (Yl) is u2 and v2. Therefore, the threshold value H, the first color components y1, u1, v1 and the second color components y2, u2, v2 are calculated by the following formula 2. In Equation 2, p is the number of Yh and q is Yl.
And p + q = 16 (= n × m). Also, U
(Yh) and V (Yh) have a luminance component of Yh (bm
Is the value of the color difference component for the pixel,
The luminance components of U (Yl) and V (Yl) take Yl (b
It is the value of the color difference component for a pixel where m takes "0".

[0019]

H = [(ΣYi) −Ymax−Ymin] / 14 y1 = (ΣYh) / p u1 = (ΣU (Yh)) / p v1 = (ΣV (Yh)) / p y2 = (ΣYl) / q u2 = (ΣU (Yl)) / q v2 = (ΣV (Yl)) / q

FIG. 4 shows a concrete calculation of the threshold value H, the luminance components y1 and y2, and the color difference components u1, u2, v1 and v2. As shown in FIG. 4, it is assumed that the intra-block pixel luminance components Y1 to Y16 have values as shown in FIG. In this case, the maximum value Ymax of the luminance component is the value “10” of Y9, and the minimum value Ymin is Y.
The value of 16 is “1”. Therefore, when the average value of the other luminance components except Y9 and Y16 is calculated, (83-10-1)
/14=5.14. Therefore, the threshold value H = 5.14.

The luminance component Yh having a value equal to or larger than the threshold value H is represented by Y1, Y5, Y as shown by the hatched portion in FIG. 4 (b).
Seven of 6, Y9, Y10, Y13, and Y14 correspond, and from the average value thereof, y1 = 77.1. Also, for the luminance component Yl smaller than the threshold value H, Y
2, Y3, Y4, Y7, Y8, Y11, Y12, Y1
9 corresponding to 5 and Y16 correspond to y2 =
It becomes 3.22. Then, as shown in FIG.
The number of Yh values is uniformly regarded as the block luminance component y1 and is represented by the code “1”. On the other hand, the nine Yl values are uniformly treated as the block luminance component y2 and is represented by the code “0”.
Create a binarized bm. That is, bm = (“10
Similarly, u1 and v1 are calculated from the color difference component average values of the pixels having the luminance component Yh, and u2 and v2 are calculated from the color difference component average values of the pixels having the luminance component Yl. Figure 4
As shown in (d), a block consisting of 16 pixels includes a color represented by a first color component y1, u1, v1 represented by a shaded portion and a second color component y2, u2, v2 represented by a dot portion.
Will be represented by two colors.

The block encoding unit 12 supplies the block component composed of the block color components y1, u1, v1, y2, u2, v2 and bm generated by the encoding unit 122 to the color reduction unit 15. Color reduction part 15
Then, CL of all block components for one frame supplied
The data is stored in a frame buffer (not shown) of the UT creation unit 151 to create a CLUT. That is, the CLUT creating unit 1
51 stored in the frame buffer (160 × 1
20 blocks) × 2 colors = 38400 sets of all color components yx,
From uy, vz, in the order that the characteristics of the image are well represented, for example,
The 56 color reduction data y, u, v are determined. The subtraction data y, u, and v are determined by, for example, the median division method (median cu
t) and the frequency equalization method (population equalize) are used to determine the data that well represent the features of the image. Then, as shown in FIG. 5A, each color reduction data y, u,
"00" to "FF" as an index for v
A CLUT is created by associating the indexes with the above, and is stored in a predetermined storage area. Also, the created C
The LUT is supplied to the transmission unit 14 and the reception unit 24 on the reception side.
Sent to. Further, in the CLUT creation unit 151, CL
In the UT creation process, the color-reduction block components Cx, Cy, and bm having the highest frequency are extracted and supplied to the reference block buffer 132 of the compressing unit 13.

Subsequently, in the color reduction processing section 152, for each block stored in the frame buffer, the first color components y1, u1, v1 and the second color components y2, u2, v2
The color reduction processing is performed on the two colors. That is, the color that most approximates each component is selected from the CLUT and converted into the corresponding indexes C1 and C2. For example, if the luminance components y of the first color components y1, u1, v1 and the color difference components u, v are yl, um, vn that are the same or most similar, the corresponding C1 is selected. Also, the second color component y
2, u2, v2 are the same or most similar to y
For p, uq, vr, the corresponding C2 is selected. As a result, the block data A (FIG. 4D) generated by the encoding unit 122 becomes as shown in FIG. 5B.
The format of the color-reduction block data B after color-reduction at this stage is such that the indexes C1 and C2 are each 8 bits and the bitmap bm is 16 bits, as shown in FIG. The C1, C2, and bm are sequentially supplied to the line buffer 131 of the compressing unit 13 as the components of the subtractive color block data B for each block.

The compressing section 13 sequentially stores the subtractive color block data B supplied from the block encoding section 12 in the line buffer 131. Then, the encoding unit 13
3, the reduced color block data B (C1, C2, b) of the processing block Cb (latest block) to be encoded
For m), data encoding is performed between the reference block (Rb) and the peripheral block of the processing block Cb according to the block component matching condition.

FIG. 6 is for explaining the concept of data coding under the block component matching condition. In this figure, a case will be described in which a block indicated by diagonal lines in a frame divided into a size of 4 pixels × 4 pixels is encoded as a processing block Cb. Here, in the line buffer 131, the block that has been encoded before the processing block Cb is Pb. Further, the upper left (2nd row and 2nd column) of the processing block Cb processed one column before is P
rl, Pr directly above (3 rows and 2 columns) Pr, top right (4 rows and 2 columns) P
rr.

Then, the subtractive color block component of the processing block Cb and these comparison blocks Rb, Pb, Prl, P
The block components of r and Prr are compared, a matching block is searched for, and according to the encoding rule shown in FIG.
Encode with matching blocks. That is, when the processing block Cb matches the reference block Rb, the block components C1, C2, bm of Cb (total 3
2 bits) are compressed into 2 bits of "00". Of the components C1, C2, and bm of the subtractive color block data B, C1 and C2 are completely matched to perform reversible compression, and bm is matched if they are similar within a certain range. It is supposed to be judged as.

On the other hand, when there is no matching block in the comparison block, the coding unit 133 determines that the block component of the processing block Cb is any of the reference block Rb, the block immediately above Pr, and the immediately preceding block Pb. It is determined whether or not they are the most similar, and coding is performed using similar blocks according to FIG. In other words, a 2-digit or 3-digit header indicating the most similar block and 3 representing the matching / mismatching of the color-reduction block components C1, C2, and bm.
It is represented by a bit map (“●●●”) and a component (★★ ...) Of the processing block Cb that does not match. Here, "●●
"" Is a bit string consisting of 0 or 1, and is "0" when they match and "1" when they do not match. Also, * is a numerical value indicating the actual component value of the processing block Cb, and each component value is 8 Bits (C1 and C2) or 16 bits (b
It is represented by m).

For example, it is assumed that the values of both components of the immediately preceding block Pb similar to the processing block Cb are as follows. Cb; C1 = 0b, C2 = 11, bm = x Pb; Cx = 0b, Cy = 12, bm = x In this case, the code of the processing block Cb is as follows. That is, the header is represented by the bit string “111” indicating the immediately preceding block which is a similar block. In the match / mismatch bitmap, C1 and bm match, and C2 does not match, so the bit string “0” is set in the order of C1, C2, and bm.
10 ”. The mismatch component value of C2 is represented by the value C2 = 11 of the processing block Cb. Therefore, the code of the processing block Cb is the header (“ 111 ”) 3 bits + match / mismatch bitmap. (“010”) 3 bits + unmatched component value (11) 8 bits.

The coded data of the color-reduction block component of the processing block Cb coded by the compressor 13 has the minimum data amount when it matches the reference block Rb. In this case, the data amount of one block at the time of division by the division unit 121 is 8 × 3 × 16 = 384 bits, which means that the data has been compressed to 1/192. On the other hand, the maximum encoded data by the compressor 13 is that there is no frame-to-frame match and all components of the processing block Cb are
This is the case when they do not match with all the components of Rb, Pr, and Pb.
In this case, if the encoding is performed based on the reference block Rb, the amount of encoded data in the compressing unit 13 is 2 + 3 + 32 = 3.
7 bits, and the compression rate in this case is 37/384 = 1
It becomes /10.38.

By the above processing, the encoded data of the subtractive color block data B encoded by the encoding unit 133 is supplied from the compressing unit 13 to the transmitting unit 14 for each block. In the transmission unit 14, the CLUT, the reference block Rb,
Then, the coded data of the color-reduction component B is transmitted by various communication means such as a telephone network, a dedicated line, wireless communication, ISDN (Integrated Service Digital Network), computer communication, etc.
It is transmitted to the receiving side shown in FIG. Receiver 2 on the receiving side
When receiving the data transmitted from the transmission side, the data No. 4 supplies the data to the separating unit 234 of the decompressing unit 23. Separation unit 23
In 4, the CLUT, the reference block Rb, and the encoded data to be received thereafter are separated from the supplied received data, the CLUT is supplied to the output unit 21, and the reference block R
b is stored in the reference block buffer 233, and the encoded data is supplied to the decoding unit 231.

The decoding unit 231 compares the coded data among the reference block Rb, the immediately preceding block Pb, the upper left block Prl, and the block immediately above Pr according to the coded code allocation shown in FIGS. 7 (a) and 7 (b). Select the block to be used. Then, if the encoded data is a code indicating a match between blocks, the block components C1, C2, and bm stored in the corresponding position (which have already been decoded) are stored in the processing block Db that is the current processing target. It is stored in the line buffer 232 as a block component. If the coded data is a code indicating a mismatch between blocks, FIG.
According to the encoding rule of (b), a block component of the processing block Db is generated from the block component of the similar block designated by the header part of the encoded data and the encoded data,
It is stored in the line buffer 232.

When the pre-encoding block components C1, C2, and bm of the processing block Db are generated by the decoding unit 231, they are supplied to the block decoding unit 22. The block decoding unit 22 generates 4 × 4 pixel data (pixel value for index color display) from the block components C1, C2, and bm. That is, if the color-reduction block component bm is the bitmap shown in FIG. 8A, the pixel whose bitmap data is “0” is C1 and is “1” as shown in FIG. 8B. The pixel becomes C2. In this way, the decoded block data of 4 × 4 = 16 pixels becomes data indicating two types of index colors C1 and C2 as a whole.

C1 indicating this index color and C
The index value of 2 is supplied to the output unit 21 for each pixel. The output unit 21 converts the CLUT separated and supplied by the separation unit 234 into the corresponding luminance signal y and color difference signals u and v, further D / A-converted, and then videophone, videoconference, and communication karaoke not shown. It is displayed on various display devices such as CRT, plasma display, liquid crystal display device, etc. Here, the output unit 21 determines that the connected display device is not the y, u, v data but the RG.
When B data is requested, R, G, and B data are calculated and output from y, u, and v data according to the following formula 3.

[0034]

B = 1.773u + y R = 1.403v + y G = y− (0.564 / 0.587) u + (0.713
/0.587)v

As described above, according to the embodiment, the data is already compressed to ⅙ for each block by the block encoding unit 12 before being encoded by the compressing unit 13. The buffer size of 131 can be reduced. Further, the block component encoding process by the encoding unit 133 and the encoded data decoding process by the decoding unit 232 are only the detection of interframe coincidence and the conversion according to the encoding rule of FIG. Even if there is no special hardware that processes the decoding at high speed, the software can handle it sufficiently. Therefore, in the case of an image compression communication system in which a still image or a moving image is transmitted from one place such as a central station in a certain building and is received at a plurality of places, both the transmitting side device and the receiving side device are used. Image processing can be performed by using a small buffer and by software processing. As described above, according to the present embodiment, the image compression communication system can have a simple configuration. Also,
In this embodiment, since the CLUT is created on the transmitting side,
Since it is not necessary for the receiving side device to perform the color reduction processing, the processing speed can be increased. Furthermore, since the color reduction processing is performed on the block data, the color reduction processing in the compression device can be performed at high speed.

In the embodiment described above, the bm of the block component is judged to be a match if they are similar within a certain range, but the present invention is not limited to this and it is a perfect match. You may use the case that you are doing. Also, instead of an exact match, for example, all "1",
64 typical bitmap tables are created in advance, such as "0" for all, "1" for upper half, "1" for right half, "1" for every other column, and so on. May be specified by a 6-bit code. In this way, since the bm of the block component is compared by each of the 64 tables represented by 6 bits, the degree of coincidence between blocks is increased and the data is further compressed. Even if they do not match, a 16-bit bitmap is represented by a 6-bit code string, so 10-bit data is compressed.

In the present embodiment, the bitmap b
Although m is represented by 16 bits, it may be represented by dividing it into upper 8 bits and lower 8 bits. That is, the components of the color-reduction block data B are C1 (8 bits), C2 (8 bits),
It may be represented by bmh (8 bits) and bml (8 bits), and data encoding may be performed according to the block component matching condition with the peripheral blocks of the reference block Rb and the processing block Cb. In this case, the interblock match code is the same as in FIG. 7A, and the interblock component match code is Rb.
“10 + ●●●●● + ★…”, Pr “11”
0 + ●●●● + ★… ”, Pb“ 111 + ●●●●●
+ ★… ”

Further, in this embodiment, each block component y
Each of 1, u1, v1, y2, u2, and v2 is represented by 8 bits, but the present invention is not limited to this.
Each block component is represented by 6 bits, and 5 bits and 4
It may be represented by another number of bits such as bits. further,
y1, y1 is 8 bits or 6 bits, u1, u2, v
The luminance signal and the color difference signal may be represented by different numbers of bits such that 1 and v2 are 6 bits or 5 bits.

Further, when the block color difference components u and v are calculated from the color difference components U and V, in the embodiment described above,
Although the average value of U and V is used, the values of the color components U and V that have the highest frequency in the processing block may be used as the block color components u and v. Also, the two block luminance components y
As the threshold value H for distinguishing 1 and y2, the luminance component Ym
Although the average value of values excluding ax and Ymin is used, a simple average value of all luminance components Y may be used as the threshold value H,
Further, an arbitrary value may be set according to the type of image data to be compressed.

Further, in this embodiment, all the block components of the processing block Cb are Rb, Pb, Prl, Pr, prr.
7B, as shown in FIG. 7B, Rb, Pf, and Pb are selected as similar block selection targets.
However, the present invention is not limited to this, and an arbitrary block may be determined in advance and a similar block may be selected from it. For example, as selection targets, Rb and Pr
The above two cases, or the three cases of Rb, Pr, and Pb may be used. Further, in this embodiment, the processing block C
Although Rb, Pb, Prl, Pr, and prr are used as targets for determining whether or not all block components with b match, in the present invention, in addition, Rb with a high matching frequency,
It may be three targets of Pb and Pr. By reducing the number of comparison objects, the processing by software can be further facilitated.

Further, in the present embodiment, the CLUT is created by selecting 256 colors that are frequently used from the full colors.
The present invention is not limited to this, and for example, 128 colors, 64
The number of colors may be reduced to any other number such as 512 colors, 1024 colors, and the like. Furthermore, in the present embodiment, each block component y, u, v is represented by 8 bits, but it may be represented by 6 bits or 5 bits.

Further, in this embodiment, RGB data and YU are used.
Although the formulas 1 and 3 are used as the conversion formulas for V data, other conversion formulas, for example, the formulas 4 and 5 may be used in the present invention.

[0042]

## EQU00004 ## Y = 0.299R + 0.587G + 0.114B U = B-Y V = R-Y

[0043]

## EQU5 ## G = Y- (0.299 / 0.587) V-
(0.114 / 0.587) U B = Y + U R = Y + V

Further, in the embodiment, when the reference block Rb is determined, the most frequent color-reduction block data is detected at the time of creating the CLUT, and this is used as the reference block Rb. However, the present invention is not limited to this. Alternatively, it may be determined in consideration of the compression ratio after compression in the compression unit 13. For example, not only the most frequently used color reduction block data but also a plurality of higher color reduction block data Rb
, Rb2, Rb3, ... Are extracted as reference blocks. Then, the compressing unit 13 encodes each reference block, and determines the subtractive color block data with the smallest data amount after encoding as the reference block. In this case, the encoding unit 133 determines the reference block having the smallest data amount by counting the data amount after encoding each reference block Rb1, Rb2, ....

Further, a frame buffer is arranged between the color-reduction section 15 and the compression section 13 to store the color-reduction block data B for one frame. And the compressing section 13
In, the coding may be performed for one frame without the reference block Rb, and the block most used as the interblock match may be adopted as the reference block Rb. Further, the color-reduction block data with the highest frequency may be used as a temporary reference block Rb ′, one frame of coding may be performed, and the block most used as the interblock match may be used as the reference block Rb. In this case, the temporary reference block R
If b'is used most as a block match, it is used as it is as a reference block. The line buffer 131 may be used as a frame buffer instead of being arranged between the color reduction unit 15 and the compressing unit 13.

[0046]

According to the present invention, each pixel data obtains image data composed of a luminance signal and a color difference signal, and n
N × showing the distribution of the first color component and the second color component having the luminance signal and the color difference signal, and the distribution of the first color component and the second color component, which represents the generated block A CLUT for compressing data by generating a block component composed of a bit map of m and performing color reduction processing on the first color component and the second color component to an arbitrary number of colors
Is generated and transmitted together with the block component, it is possible to perform image compression that can be easily expanded by the device on the image expansion side.

[Brief description of drawings]

FIG. 1 is a system configuration diagram of an image compression communication system to which an image compression apparatus according to an embodiment of the present invention is applied.

FIG. 2 is an explanatory diagram of a case where image data is divided into blocks each having 4 pixels × 4 pixels.

FIG. 3 is the same as the above, from the YUV data in the encoding unit, the block luminance components y1 and y2, the block color difference component u
It is explanatory drawing at the time of generating 1, u2, v1, v2, and bm.

FIG. 4 is the same as the above, in the encoding unit, a threshold value H, luminance components y1, y2, and color difference components u1, u2, v1, v2.
It is explanatory drawing about the concrete calculation of.

FIG. 5 is an explanatory diagram showing a CLUT created by the color reduction unit and a data distribution after color reduction.

FIG. 6 is a conceptual explanatory diagram of data encoding according to a block component matching condition in the encoding unit.

FIG. 7 is a diagram showing an encoding rule of block components in the encoding unit.

FIG. 8 is an explanatory diagram for generating index color data of 4 × 4 pixels from the subtractive color block components C1, C2, and bm in the block decoding unit.

[Explanation of symbols]

 11 YUV acquisition unit 12 block encoding unit 121 division unit 122 encoding unit 13 compression unit 131 line buffer 132 reference block buffer 133 encoding unit 14 transmission unit 15 color reduction unit 151 CLUT creation unit 152 color reduction processing unit 21 output unit 22 block decoding unit 23 Decompressor 231 Decoder 232 Line Buffer 233 Reference Block Buffer 24 Receiver

Claims (4)

[Claims] 1. Image data acquisition means for acquiring image data in which each pixel data is composed of a luminance signal and a color difference signal, and image data acquired by the image data acquisition means,
Block generating means for generating block data composed of n × m pixels, and first and second color components having a luminance signal and a color difference signal, which represent the block from the block data generated by the block generating means. , And an encoding means for generating a block component consisting of an n × m bitmap showing the distribution of the first color component and the second color component, and the first color component and the second color component generated by this encoding means.
CL for performing color reduction processing on any number of color components
An image compression apparatus comprising: a CLUT creating unit that creates a UT; and a sending unit that sends the CLUT created by the CLUT creating unit and the block component created by the encoding unit. 2. The CLU for the first color component and the second color component of the block component generated by the encoding means.
Color reduction processing is performed by the CLUT created by the T creation means,
A color reduction unit for generating a color-reduced first color component, a second color component, and a block component consisting of an n × m bitmap, wherein the transmission unit is a CLU created by a CLUT creation unit.
The image compression apparatus according to claim 1, wherein T and the block component after the color reduction generated by the color reduction unit are transmitted. 3. The block component generated by the encoding means or the block component after color reduction generated by the color reduction means is selected as a similar block whose block component is the same or most similar to the block component. And a coding unit that codes from the code data indicating the similar block and the difference data with respect to the block component of the similar block, and the transmitting unit includes the CLU created by the CLUT creating unit.
The image compression apparatus according to claim 1 or 2, wherein encoded data that has been encoded by T and the encoding means is transmitted. 4. The CLU created by the CLUT creating means
The image data acquisition unit includes a multi-frequency block detection unit that detects a block having a high frequency of color reduction block components after the color reduction processing by T, the image data acquisition unit acquires image data of one frame for a still image, and the encoding unit is The block of the block detected by the multi-frequency block detection means is further selected by selecting a block that matches or includes the block component of the block detected by the multi-frequency block detection means. An image compression apparatus according to any one of claims 1 to 3, which transmits a component.