JPH07147638A - Image processor - Google Patents

Abstract

PURPOSE:To reduce communication time and line use rate in the case of trans mission with a communication line by effectively compressing color binary images and improving the use efficiency of a storage device. CONSTITUTION:Concerning the image processor for compressing/extending color binary image data expressed by M color planes, this device is provided with a color plane converting means 104 for reversibly converting M color planes and N (>M) color planes by generating new color planes by performing logical arithmetic processing among the plural color planes and a compressing/ extending means 105 for performing reversible compression/extension for each of N respective color planes, and the color plane converting means 104 detects the representative color binary image data and changes the contents of logical arithmetic processing corresponding to the detected color binary image data.

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 efficiently encoding / decoding binary color image data, and more particularly to reading an original using an image information input device such as a scanner and binarizing the original. The present invention relates to a technique effectively applied to a technique of performing compression processing after processing, recording encoded data in a storage device, or transmitting it via a communication line.

[0002]

2. Description of the Related Art For example, an A4 size original is 400 dpi.
When the image information is read by an image input device such as a scanner at a density of, the image information has a data amount of about 16 MB, assuming that each pixel has a density of 256 gradations. If this is a color original, 16MB for each RGB color, for a total of 48
It is a huge amount of data in MB. When storing a huge amount of image information in a storage device or transmitting it through a communication line, it is inefficient to handle it as it is.
A process for reducing the amount of data, that is, a compression process is required. There are two types of compression processing: lossless compression and lossy compression. The former retains the original image data completely even after restoration, and the latter makes the restored image as close to the original image as possible. A compression process that can be reproduced is desirable.

Further, a method of reducing the data amount by reducing the number of gradations for each pixel by a method of maintaining the gradation and the density distribution of the original image as much as possible can be considered. As such a general method, the gradation of pixels is turned on / off by a threshold value.
There is a valuation method, and many such methods have been proposed.
It is known that pseudo halftone processing such as dithering and error diffusion is effective when an original image is effectively reproduced by a binary image. This also applies to a color binary image in which each pixel of a plurality of color planes (for example, RGB, YMC, etc.) is represented by 1 bit. The color pseudo-halftone image obtained by subjecting each of the plurality of color planes to the pseudo-halftone processing can relatively faithfully reproduce the original image with a small amount of data. Moreover, the color binary image can improve the fidelity with respect to the original image by expanding the number of color planes constituting the color binary image, and can reproduce a clearer image. For example, if there are three color planes, it is possible to reproduce 8 colors for each pixel by turning on / off each plane. However, by using 5 color planes, 32 colors are provided for each pixel, and 256 colors are used for 8 pixels. It is possible to extend such as reproduction of. As described above, the number of color planes constituting the color binary image is determined according to the required degree of faithful reproduction with respect to the original image.

Further, since the printing apparatus only turns on / off the pixels, the apparatus can be simplified, which is an effective method for a small color FAX, a color printer or the like. When the data amount of the color pseudo-halftone image is compared with the data amount of the monochrome image used for FAX or the like, the data amount is several times as large, so that when it is stored in the storage device, the writing time and the file capacity become large. In addition, when data is transmitted via a communication line, communication time, line utilization rate, etc. are also large. Therefore, it is necessary to efficiently compress the color pseudo halftone image.

When compared with the original image in terms of image quality, the color pseudo-halftone image is inevitable to some extent when it is a binary image in which the gradation of its pixels is only ON / OFF. Therefore, when compressing this color pseudo-halftone image, it is necessary to use lossless compression in order to prevent further deterioration than lossy compression in which the original image is not completely restored.

[0006]

However, it is known that the reversible compression used in FAX or the like cannot effectively compress pseudo-halftone images. For example, the Huffman coding used in FAX compresses the data amount by replacing the continuous length of pixels of the same color, that is, the run length and the difference from the previous scan data with a code. Pseudo-halftone images reproduce gradation by turning pixels on and off, so when Huffman coding is used, the probability of occurrence of short run lengths is high, so the number of codes increases and the code amount increases. And the compression ratio becomes worse. Even in pseudo-halftone, the dither image has a periodic pixel arrangement, so the predictive coding method is effective even in the lossless compression.
It is known that lossless compression does not work for error diffusion images that do not have periodicity.

This also applies to color pseudo-halftone images. In the color pseudo-halftone image, each of the plurality of planes is a pseudo-halftone image, and in order to compress the plane, the compression processing is performed for each of the plurality of planes. For this reason as well, lossless compression does not work for the same reason as described above, and when saving this color pseudo halftone image in a storage device or transmitting it via a communication line, the data must be handled in a state of being almost uncompressed. was there.

The present invention has been made to solve the above problems, and an object of the present invention is to effectively compress a color binary image, improve the utilization efficiency of a storage device, and transmit it via a communication line. An object of the present invention is to provide an image processing device capable of reducing communication time and line usage rate.

[0009]

To this end, the present invention provides M
An image processing apparatus for compressing / decompressing color binary image data represented by a plurality of color planes, wherein a new color plane is generated by performing logical operation processing between a plurality of color planes. Color plane conversion means for reversibly converting N color planes and N (> M) color planes, and compression / expansion means for performing reversible compression / expansion for each of the N color planes. The color plane conversion means detects representative color binary image data and detects the detected color 2
The content of the logical operation processing is changed according to the value image data.

[0010]

In the image processing apparatus of the present invention, the original is read in full color by using the image input means, and the binary image is generated from the read color image information by the binarizing means. Then, the color plane conversion means performs reversible color plane conversion processing on a plurality of color planes forming this color binary image, a new color plane is generated, and the compression / decompression means performs reversible compression processing. Then, the coded data is recorded in the storage means or transmitted through the communication line. Further, the coded data is decoded as necessary, and the image is output by an image output means such as a printer. In this way, before the lossless compression process, a step of performing a reversible color plane conversion process on a color binary image composed of a plurality of color planes is provided, so that the information amount of the color binary image, that is, the code amount. Can be reduced as compared with the conventional method, and storage efficiency, transmission efficiency, etc. can be improved.

The color plane conversion process is a process of comparing and referring to a plurality of color planes for each pixel, performing a logical operation process between the planes, and generating the operation process result as a new plane. The number of new color planes after processing is greater than the number of color planes forming the image. Further, since this process is reversible, it is possible to restore the original color plane from the converted color plane. Then, the lossless compression process is performed on each of the converted color planes.

For example, when the image information of a color document is read as three color planes of RGB from an image input device such as a color scanner, first, the RGB planes are binarized such as dither and error diffusion. Perform processing. Pixels at the same position in the binarized RGB plane are referred to between the planes, and logical operation processing as shown in FIG. 7 is performed. Assuming that the logical operation is “AND (logical product)”, the logical operation result plane W is turned “on” when all of the RGB are “on”, and the RGB pixel at that position is turned “off”. Generate the'G'B 'plane. In other cases, the logical operation result plane W is set to “off”, and the R′G′B ′ plane remains RGB.
The contents of the plane. Further, even when the logical operation is “OR (logical sum)” or “EXOR (exclusive logical sum)”, the arithmetic processing result is set in the logical operation result plane W and R′G ′ is set.
Similar processing is performed on the B ′ plane. FIG. 7 shows an example of these logical operation processes. Such a process is performed on all pixels to form four new color planes R'G'B'W from the three color planes of RGB. Lossless compression is performed on each of the new color planes.

When it is necessary to faithfully reproduce an image, the number of color planes forming a color binary image is not three but more than three as described above. For example, when it is necessary to reproduce 256 colors for each pixel, the image information is read as 8 color planes from an image input device such as a color scanner, and 2 color dithers, error diffusion, etc. are performed for each color plane. Perform value conversion processing. This 2
The pixel of the binarized plane is referred to, and a new color plane is generated by the logical operation processing as described above. In the new color plane, the number of “off” pixels is increased by the logical operation processing, the white background area in the image is enlarged, and as a result, the redundancy of the image is reduced. As a result, when encoding processing such as lossless compression is performed, it is possible to effectively compress each of the original color planes as compared with the case where lossless compression is performed.

Generally, when comparing the data amount after encoding, a new color plane is compressed more than the code amount obtained by directly compressing the original color plane, although the number of color planes is increased by this process. The amount of code that is used is smaller. Further, when the coded data by the above method is decoded and a new color plane is decoded, the original color plane can be restored from this color plane by performing the above logical operation processing in reverse.
That is, by using “color plane conversion processing” + “reversible compression”, it is possible to improve the compression rate when reversibly compressing a color binary image. By improving the compression ratio of the color binary image by the above method, it is possible to improve the utilization efficiency of the storage device and reduce the communication time and the line usage rate when transmitting through the communication line.

[0015]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram for explaining one embodiment of an image processing apparatus according to the present invention. In FIG. 1, the original 101
Is read by the image input device 102 such as a color scanner. The read data is converted into binary data in which each pixel is "on" or "off" by the binarizing means 103 using dither or error diffusion for each color plane forming image information. This binary data is sent to the logical operation processing circuit 104, where a plurality of color planes are compared and referenced for each pixel, logical operation processing is performed between these planes, and the operation processing result is generated as a new plane. To do. Each of these new planes is input to the lossless compression circuit 105, compressed, and then sent to the storage medium 106 such as a hard disk to be stored therein, and decrypted as necessary. In this case, the number of color planes is determined by the required degree of faithful reproduction with respect to the original image. For example, when configuring a small system, R
A configuration with three color planes such as GB and YMC is suitable. Further, in the case of faithfully reproducing the original image, a configuration with a large number of color planes such as eight that reproduce 256 colors for each pixel is suitable.

Next, some other embodiments of the present invention will be further described. FIG. 2 is a diagram for explaining another embodiment of the image processing apparatus according to the present invention. In FIG.
The original 301 is an image input device 302 such as a color scanner.
Are read by being decomposed into the three primary colors of RGB by
Each of them is independently binarizing means 303, 3 such as error diffusion.
04 and 305 are input. Binarization means 303, 30
The data binarized into RGB each 1 bit / pixel in 4 and 305 is input to the logical operation processing circuit 306, and the logical operation processing circuit 306 inputs R'G'B 'of a new color plane.
Get W. Reversible compression / decompression circuits 307, 308, 3 for each of the new color planes R'G'B'W.
In steps 09 and 310, compression processing is performed to obtain coded data. The code data thus obtained is stored in the storage means 313 such as a hard disk via the storage medium control device 312, or transmitted to the communication line via the communication control device 311.

As described above, the logical operation processing circuit 306 performs color plane conversion by logical operation processing, and the lossless compression / expansion circuits 307, 308, 309, and 310 perform lossless compression, so that a huge amount of data can be efficiently processed. It becomes possible to compress. Therefore, the storage means 313
In this case, the storage efficiency is significantly improved, and the storage capacity can be reduced. Further, even when the obtained coded data is transmitted to the communication line via the communication control device 311, it is possible to reduce the communication time at the time of transmission and the line usage rate.

The coded data stored in the storage means 313 such as a hard disk can be restored to a color binary image if necessary. In this case, the code data read from the storage unit 313 is sent to the lossless compression / expansion circuits 307, 308, 309, and 310 for each R'G'B'W. Here, the color plane R'G'B'W is restored by decoding from the coded data, and the RGB data is restored from the R'G'B'W by the logical operation processing circuit 306. The above-described processing makes it possible to completely restore the binarized color binary image, and the restored color binary image 31
5 is an image output device 3 such as a color CRT or a color printer
It is output by 14.

Next, the contents of the calculation by the logical operation processing circuit will be described. FIG. 3 is a diagram for explaining the logical operation processing by the logical operation processing circuit shown in FIG. 2, and FIG. 4 is a diagram for explaining the relationship between the color plane conversion and its code amount. In the logical operation processing by the logical operation processing circuit shown in FIG. 2, for each RGB pixel, as shown in FIG. 3A, all RGB are “on” → W is “on”, R′G′B ′ is each “off”. , And as shown in FIG. 3B. Others → W is “off” R'G'B 'is the conversion process of RGB data as it is. As a result, when the R plane and the R ′ plane are compared, the “off” region of the R ′ plane becomes wider and the redundancy of the image is also reduced. This can be explained as follows.

When encoding a binary image, run length encoding used in FAX is often adopted.
This replaces the continuous length of white pixels / black pixels in the scanning direction with a code. When run-length coding is used for the error diffusion image obtained by the binarizing means of the present embodiment, the error diffusion image expresses the contrast of the image by black and white (“on / off”) of pixels, so that a photograph etc. If error diffusion is performed on the halftone image of, there is a high probability that only a very short run length will appear. Therefore, the code amount increases, and as a result, the compression effect does not increase. This also applies when the R-plane is directly run-length encoded.

However, when the R'plane obtained from the color plane conversion by the logical operation processing is run-length encoded as in the present embodiment, as a result of the logical operation processing, the "off" area is wider than the R plane. So
The probability that a long run length will appear will increase, and the compression rate will increase. Since the G and B planes have the same tendency, the compression rate is improved as a whole. The W plane has lower redundancy and higher compression rate than the R'G'B 'plane. Therefore, as a result, rather than the total code amount obtained by reversibly compressing each RGB,
The total code amount obtained by losslessly compressing R'G'B'W is smaller, so that the compression rate is improved as a whole.

On the contrary, the logical operation processing for decoding the coded data is performed as follows.

When W is "on" → RGB is "on" respectively. When W is "off" → RGB is R'G 'as it is.
B'data The color plane R'after processing by the above calculation processing
It is possible to restore the original color plane RGB from G'B'W.

According to the present invention, as shown in FIG. 4, a plurality of planes (C ₁ , C ₂ , ...
, C _n ) 201 is losslessly compressed, and the code amount of each code data 203 is summarized as the code amount #
1. Similarly, the planes (C ′ ₁ , C ′ ₂ , ..., C ′ _n , ...) Which constitute the color plane B after color plane conversion are similarly set.
, C ′ _m ) 202 is losslessly compressed, and the code amount of each code data 204 is summarized as a code amount # 2 (n <m), and these two code amounts # 1 and # 2 are compared. Then, the code amount # 1> the code amount # 2 tends to be satisfied.

FIG. 5 is a diagram showing still another embodiment of the image processing apparatus according to the present invention, and FIG. 6 is a diagram showing an example of a combination of representative colors of YMC and RGB.

In FIG. 5, binarizing means 503, 50
The data which has been binarized by 4 and 505 and has 1 bit / pixel for each YMC is input to the representative color detection circuit 506.
The representative color detection circuit 506 calculates the distribution / frequency of the colors included in the original from the data, detects the color that appears with the highest probability, and sets that color as the representative color of the original. Eight representative colors appear in each combination of 1 bit of YMC. FIG. 6 shows a combination of representative colors by YMC.
It is (a), and the original is converted into RGB by the image input device.
FIG. 6B shows a combination of representative colors of RGB that appear when the reading and binarization processing is performed.

The representative color detected by the representative color detecting circuit 506 as described above is sent to the logical operation processing circuit 507, and the logical operation processing circuit 507 performs logical operation processing based on the data of the representative color to obtain the representative color. Perform arithmetic processing to extract.
For example, a representative color is detected from image data of YMC each 1 bit / pixel, and the result is (Y, M, C) = (1,
When the color is G of 0, 1), the color plane Y'M'C'G is generated from the YMC color plane by the logical operation processing. The contents of the logical operation processing are as follows.

When Y = “on”, M = “off”, C = “on” → Y ′ and C ′ are “off”, M is “on” G is “on” / Otherwise → Y'M'C 'remains YMC data G is “off” As a result of this arithmetic processing, the representative color is extracted to the newly generated G plane, and Y'M'C' has YMC. The data is obtained by removing the pixel portion where G was present from the plane. As a result, when comparing the Y plane and the Y ′ plane, the “off” region of the Y ′ plane becomes wider and the redundancy of the image is also reduced, so that the probability of appearance of a long white run length increases, As a result, the compression rate is improved. This also applies to the C'plane, and the compression rate is improved in each case. Also G
Similar to the W plane of the previous embodiment, the plane has lower redundancy and higher compression rate than the Y'M'C 'plane.

This new color plane Y'M'C'G
Each of the reversible compression / decompression circuits 508, 509, 51.
The compression processing is performed at 0 and 511 to obtain coded data. The code data thus obtained is stored in the storage device 514 such as a hard disk via the storage medium controller 513.
Memorized in. Conversely, the color plane Y'M'C'G
When converting from the color plane to the color plane YMC, (YMC) = (1, 0, 1) may be set only when M is “on”.

With the above structure, the storage efficiency of the storage means 514 is significantly improved as in the previous embodiment, and the storage means 51 is improved.
4 can be reduced, and when the obtained coded data is transmitted to the communication line via the communication control device 512, the communication time and line utilization rate at the time of transmission can be reduced. Storage means 51 such as a hard disk
Similarly, the data stored in 4 can be completely restored to a color binary image if desired. As a result, by providing this representative color detection circuit, the reversible compression rate of the color binary image can be greatly improved.

FIG. 8 is a diagram showing still another embodiment of the image processing apparatus according to the present invention. When the color binary image is compressed as described above, the compression rate is improved by performing the logical operation process and then the reversible compression, but the compression rate may rarely be improved depending on the document. In order to deal with the case where such a problem occurs, in the present embodiment, the lossless compression process is performed for each of the case where the logical operation process is performed on the color binary image and the case where it is not performed, and the code amount is compared. The smaller one of the processing results is retained.

In FIG. 8, binarization means 603, 60
4 and 605, the RGB data of 1 bit / pixel is input to the logical operation processing circuit 606, and a new color plane R′G′B′W is obtained. The specific content of the logical operation processing is a method similar to the logical operation processing content of the previous embodiment. This newly obtained color plane R'G'B'W
Lossless compression / expansion circuits 607, 608, 6 for each plane
09 and 610 to obtain coded data. In the code amount calculation circuit 611, the code amount obtained by performing the logical operation process and the lossless compression process on the color binary image is calculated.

Further, the binarization processing circuits 603, 604, 6
Without performing logical operation processing on the output result from 05,
It is directly input to the lossless compression / expansion circuits 617, 618, and 619, and the code amount calculation circuit 616 calculates the code amount when the logical operation processing is not performed. In the code amount determination circuit 612, the code amount when the logical operation processing is performed and the code amount when not performed are compared to determine which one has the smaller code amount, and the determination processing result is used as the communication control device 613 and the storage medium control. Input to the device 614. Communication control device 613
Converts the input code data into a form suitable for the connected communication line and transmits it. Storage medium control device 614
Writes and stores the input code data in a form suitable for the storage unit 615 such as a hard disk.

The code data stored in the storage means 615 can be restored as necessary. In this case, the storage medium control device 614 reads the code data stored in the storage medium 615, and the code amount determination circuit 61
Enter in 2. In the code amount determination circuit 612, it is determined whether this code data is a code compressed through logical operation processing or a code compressed without logical operation processing. The determination result is input to the logical operation processing circuit 606, and it is determined whether to perform the logical operation processing at the time of decoding based on the determination result. The code data is a lossless compression / expansion circuit 6
The data is decoded by 07, 608, 609 and 610 and input to the logical operation processing circuit 606. Logical operation processing circuit 6
In 06, the logical operation processing is performed in the case of the data subjected to the logical operation processing to restore the original color plane, and in the case of the data not undergoing the logical operation processing, the original color plane is restored at the decompressed stage. I do nothing.
By the above arithmetic processing, the color binary image 621 after binarization
Can be completely restored, as in the previous embodiment.

The present invention is not limited to the above-described embodiments, and within the scope of the invention,
It is possible to change the processing content and the like. For example, when converting from M color planes to N color planes (N> M), logical operations may be performed only on a part of the M color planes and not on all the M color planes. The color planes may be further converted into N'sheets (N '> N) for compression and expansion.

[0036]

As is apparent from the above description, according to the present invention, a reversible color plane conversion process using a logical operation process between a plurality of color planes is applied to a color binary image having a low compression rate in the related art. Therefore, the probability that a long run length appears in each new color plane can be increased. Therefore, by performing lossless compression after the color plane conversion process, the compression rate of code data as a whole is improved, the utilization efficiency of the storage device is improved, and the communication time and line utilization rate when transmitting via a communication line are reduced. can do.

[Brief description of drawings]

FIG. 1 is a diagram for explaining one embodiment of an image processing apparatus according to the present invention.

FIG. 2 is a diagram for explaining another embodiment of the image processing apparatus according to the present invention.

FIG. 3 is a diagram for explaining logical operation processing by the logical operation processing circuit shown in FIG.

FIG. 4 is a diagram for explaining the relationship between color plane conversion and its code amount.

FIG. 5 is a diagram showing still another embodiment of the image processing apparatus according to the present invention.

FIG. 6 is a diagram showing an example of combinations of representative colors of YMC and RGB.

FIG. 7 is a diagram illustrating an example of a logical operation process.

FIG. 8 is a diagram showing still another embodiment of the image processing apparatus according to the present invention.

[Explanation of symbols]

Reference numeral 101 ... Original document, 102 ... Image input device, 103 ... Binarizing means, 104 ... Logical operation processing circuit, 105 ... Reversible compression means, 106 ... Recording medium

Claims (2)

[Claims] 1. An image processing apparatus for compressing / expanding color binary image data represented by M color planes, wherein a new color plane is generated by performing logical operation processing between a plurality of color planes. As a result, a color plane conversion unit that reversibly converts the M color planes and the N (> M) color planes, and a compression / decompression unit that performs reversible compression / decompression for each of the N color planes. An image processing apparatus comprising: 2. The color plane conversion means detects a representative color binary image data, and detects the detected color 2
The image processing apparatus according to claim 1, wherein the content of the logical operation processing is changed according to the value image data.