JPH04142879A - Picture coding and decoding system - Google Patents

Abstract

PURPOSE:To improve the compression efficiency by coding a luminance component and a chrominance component with respect to a halftone area and coding the luminance component with respect to a character area. CONSTITUTION:An inputted primary color signal is binarized by a binarizing section 1, a base color detection section 2 detects a base color, a color space conversion section 3 separates the inputted primary color signal into a luminance component and a chrominance component and a color signal generating section 5 uses background color information from the background color detection section 2 to generate color signals Cr1, Cb1. As the result of discrimination by an area discrimination section 6, when a block is resident in a halftone area, a halftone coding section 7 is operated to code the luminance signal and chrominance signal Cr1, Cb1 in the unit of blocks. When a block is in a character area, a character area coding section 8 is active to allow a luminance signal binarizing section 4 to code the binarized luminance signal in the unit of blocks. Thus, a high compression rate is obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for encoding and reconstructing color images, and in particular to image encoding for encoding an image of an original with a colored background and the encoding method. The present invention relates to a decoding method for decoding a converted image pig.

[Prior Art] Color images have a much larger amount of data than conventional monochrome binary images, and when storing or transmitting color document images, encoding processing is performed to reduce the amount of data. Data needs to be compressed. In order to obtain a high compression rate, it is necessary to use an encoding method suitable for the characteristics of each image, and various encoding methods have already been proposed for this purpose.

In particular, for documents containing a mixture of text and photographs, the compression ratio is increased by dividing the image into a color image part and a monochrome image part and adaptively encoding them.

For example, in the technique described in Japanese Patent Application Laid-Open No. 62-1373, an input image is separated into a luminance signal and a chromaticity signal, and it is determined whether the input image is a color image from the amount of the separated chromaticity signals. However, when producing a color image, the luminance signal and chromaticity signal are encoded, and when the image is not a color image, only the luminance signal is encoded to improve compression efficiency.

[Problems to be Solved by the Invention] However, the conventional encoding method described in 1. is effective for documents containing a mixture of black and white text and color photographs, but it is effective for documents with a colored background. Since all images are processed as color images, there is a problem in that a high compression ratio cannot be obtained.

The purpose of the present invention is to separate a document into a character area and a halftone area by detecting the background color of the original, encode only the luminance component for black characters in the character area, and encode the luminance component and color component for color characters. , An object of the present invention is to provide an image encoding and decoding method for encoding a luminance component and a color component in a halftone region in the same way as in the past, and decoding the encoded image data.

[Means for Solving the Problems] In order to achieve the above object, in the invention according to claim (1), in an image encoding method that encodes a color image for each predetermined area, the color of the base of the document is changed. means for detecting, means for correcting a color component value of an input image signal according to the detected background color, and determining whether a predetermined area is a halftone area or a character area based on the corrected color component value. and means for encoding a luminance component and a color component for a halftone area, and encoding a luminance component for a character area, in response to the determination result of the determination means. It is characterized by the fact that it is equipped with

In the invention described in claim (2), in an image decoding method for decoding a color image for each predetermined area, a first means for decoding a base color signal from encoded image data;
a second means for decoding area information specifying whether a predetermined area is a halftone area or a character area; and a decoding process corresponding to the predetermined area in response to the output of the second means. The present invention is characterized in that it comprises means for selecting, and means for reproducing color signals in response to the outputs of the first and second means.

In the invention described in claim (3), in an image encoding method for encoding a color image for each predetermined area, there is provided a means for detecting the background color of one document, and an input image according to the detected background color. means for correcting the value of a color component of the signal; first means for determining whether the area is a predetermined area, a halftone area, or a character area from the corrected value of the color component;
a second means for detecting black characters from the luminance component and the color component;
means for encoding a luminance component and a color component for a halftone region in response to the determination result of the first means; and a means for encoding a luminance component and a color component for a black text region in response to the output of the second means; and a means for encoding the binarized luminance component and color component for the color character area in response to the outputs of the first and second means described in iii. It is characterized by

In the invention described in claim (4), in an image decoding method for decoding a color image for each predetermined area, a first means for decoding a base color signal from encoded image data;
a second means for decoding area information specifying whether a predetermined area is a halftone area, a black text area, or a color text area; and responding to the output of the second means to correspond to the predetermined area. means for selecting a decoding process, and the first and second decoding processes.
and means for reproducing a color signal in response to the output of the means.

The invention according to claim (5) is characterized in that in order to store or transmit color image data, the color image data is composed of base color information, area information, and encoded data.

The invention as set forth in claim (6) is characterized in that the background color detection means includes means for accumulating human images for the time required to detect the color of the background of the document.

[For production]

According to the above-described means, the input primary color signals are binarized by the binarization section, and the binarized primary color signals R1, Gl, Bl
is input to the background color detection section, and in the background color detection section, the signal R
1. Create a histogram for each Gl and Bl. The background color is detected by comparing this hiss i-gram with the total number of pixels of the input plurality of lines, and the background color information is sent to the color signal generation section. In addition, the color space conversion section separates the six-dimensional primary color signal into a luminance component and a color component, and this color signal is sent to the color signal generation section. Color signals Crl, Cbl using information
generate. This color signal Cry.

Cl) l is sent to the area determination section, and the area determination section determines whether the l block is a halftone area or a character area. As a result of the determination, if the block is a halftone area, the halftone encoding unit operates and encodes the luminance signal and color signals Crl and Cbl in block units, and if the block is a text area, it is coded as a text area code. The converting section operates to encode the luminance signal binarized by the luminance signal binarizing section in units of blocks.

[Example] Hereinafter, an example of the present invention will be specifically described using the drawings.

FIG. 1 is a block diagram of a color image encoding method according to a first embodiment of the present invention. In the figure, 1 is the primary color belief r;! of a color image input from the outside. , G.
A binarization unit binarizes 11 using a predetermined threshold value, and 2 is a binarized primary color signal R1, Gl.

A base color detection unit 3 detects the base color from B1 and converts the primary color signals R, G, and B into a luminance signal Y and a color signal Cr.

4 is a luminance signal binarization unit that binarizes the luminance signal Y; 5 is a color signal generation unit that generates color signals Crl and Cbl based on the signal from the base color detection unit 2; ,
6 is a generated color signal CRL.

A region determination section 7 determines a character region and a halftone region from Cbl, and a halftone encoding section 8 encodes a halftone luminance signal Y and color signals Crl and Cbl according to the region information from the region determination section 6. is a character area encoding unit that encodes the luminance signal Y1 binarized by the luminance signal binarization unit 4. Then, the background color information from the background color detection section 2 is outputted to the line 9, the halftone color information from the halftone encoding section 7 is outputted to the line 10, and the area information is outputted to the lines 11 and 12, respectively. , character data is output, and these pigs are configured in the format described below and transmitted or stored.

The present invention constructed as follows will be described in detail. In this embodiment, the background color is detected. Signals of a plurality of lines of a document that have been read in advance (for example, a margin portion of the back of the document whose background color is relatively easy to detect) are input from a buffer memory (not shown). These input primary color signals R and G.

B is binarized by the binarization unit 1, and the binarized primary color signal R
1, Gl, and Bl are input to the base color detection section 2.

The background color detection unit 2 generates a histogram +((Rl), l((Gl), l-I) for each signal R1, Gl, and Bl.
Create (B1). This histogram 1-((R1
).

The background color is detected by comparing 1-T (CI), 1-I (Bi) and the total number of pixels J of the plurality of lines that have been irritated. That is, j f H(X) /J>threshold T H]]hen
5(X) else R1.5(X)=0. The combination of Gl and Bl colors is used as the base color. The background color information S (R1
), S (Gl), and S (+31) are sent to the color signal generation section 5.

On the other hand, the color space conversion unit 3 inputs manually generated primary color signals R, G,
For example, B is a luminance component Y and a color component Cr as shown in the following equation,
Decomposes into Cb.

Y=0.3R+0.59G+0.11 BCr=0.5
1R-0,43G-0,08BCb=-0,17R-0
, 34G+0.51B This separated color signal is divided into blocks of, for example, 8×8 pixels, and among these, the color signal Cr
, Cb are sent to the color signal generation section 5. The color signal generation section 5 uses the background color information S (X) from the background color detection section 2 to perform calculations as shown in the following equation, and generates a color signal C1°Cbl.

Cr1=Cr-0,51S (R1) +〇,43S(
Gl)+o,oss (Bl) Cbl=Cb+O,]]S (R1) +〇,34S(
G I) -0,5], ] (B ]) In general, when C=aR+bG+cB, CI=C-aS
(R1)-bS (Gl)-cS (131).

These color signals C1 and Cbl are sent to the area determination unit 6, where the color signals Crl and Cbl for the previous one block are sent to the area determination unit 6.
Find the sum Sc of the absolute values of Sc, compare this sum Sc with a predetermined threshold Th2, and calculate 11 ft, +] for the block.
Determine whether there is a check area or a character area. That is, 1fSc-Σ(absolute value of Crl+absolute value of Cbl))T
h2, t h e n Halftone region else The determination signal generated by the character region region determination section 6 is sent to the halftone encoding section 7 and the character region encoding section 8, and if the block is in the halftone region, the determination signal is The tone encoding unit 7 operates and encodes the luminance signal Y and color signals Crl and Cbl in units of blocks. As this encoding method, for example, ADCT (Adaptive Di
screte Co51ne Transf.

r·n1er, applied discrete cosine transform) method is used.

On the other hand, if the block is a character area, the character area encoder 8
operates, and the luminance signal Y1 binarized by the luminance signal binarization section 4 is encoded in units of blocks.

As this encoding method, for example, M l-1 (Modi
Huffman encoding method) is used.

Data encoded in this manner is stored or transmitted.

FIG. 2 shows the encoding format. The encoded data includes base color information 21, area information 22, and encoded data 23.

However, when encoded data in the same area is consecutive, 2
The encoded data is stored continuously, omitting the area information of blocks after the block.

FIG. 3 is a block diagram of the present invention for decoding "-" encoded data.

31 is a base color information decoding unit that decodes base color information from encoded data, 32 is an area information decoding unit that decodes area information from encoded data, and 33 is a base color information decoding unit that decodes area information from encoded data. When decrypted, the n that will be sent next
When the area information decoding unit 32 decodes the encoded data for 9 blowers to reproduce the luminance signal Y, the character area decoding unit 34 decodes the coded data for 9 blowers and reproduces the luminance signal Y. Decode the encoded data for 9 minutes to obtain the color signal Crl
, Cbl and a halftone area decoding unit that reproduces the luminance signal Y, 3
5 is a color signal reproducing unit that reproduces the color signals Cr and Cb according to the decoding result of the base color information decoding unit 31; This is a color space conversion unit that restores RGB signals.

For decoding, the encoded data is first subjected to background color information 5 (R1).
, S (Gl), and S (Bl) 21 are decoded by the base color information decoding section 31. Next, the area information 22 is decoded by the area information decoding section 32, and if it is a character area, for example, an instruction is given to drive the character area decoding section 33 via a line 37.

Then, the character area decoding unit 3 continues until the next area information is input.
3 performs decoding of encoded data in the character area (1--according to the example described, MH method decoding). Then, when the next area information is read out, since the information indicates that it is a halftone area, the decoding unit 3
In response to the instruction signal from 2, the halftone region decoding section 34 performs halftone region decoding (according to the above example, ADCT method decoding).

Decoding processing is performed by repeating such processing. Then, according to the decoding result S (X) of the decoding unit 31 via the line 39, the color signal reproducing rfll 35 reproduces the color signals Cr and Cb from the decoded data as shown in the following equation. However, for the character area, Cr1=Cbl=
Subtract as O.

Cr=crl+o, 51 S (Rl) -0,4,
33(Gl)-0,083(Bl) Cb=Cbl-0,178(R1) -0,343(G
l) +0.513 (B l) In general, when C=aR + bG + cB, C =
C1+aS (R1)+bS (Gl)+cS(B1)
becomes.

The reproduced C r , C i l) and the decoded luminance signal Y are color space converted by the color space converter 36 and converted into RG
Restore the B signal.

In the first embodiment, encoding is performed after reading the image data from an external buffer and detecting the background color, but as shown in FIG. A delay buffer may be provided to store read (image) data for a time corresponding to background color detection.

In addition, as an encoding format, area information may be added for each block, and the block unit is not limited to 8 x 8 pixels, and the encoding method may also be other methods. May be adopted. Further, the conversion formula from the primary color signal to the luminance signal and color signal is not limited to the embodiment described in 1.

FIG. 5 is a block diagram of a color image encoding method according to a second embodiment of the present invention. In this second example,
1) In addition to the configuration of the first embodiment described in 1j, a black character determination unit 51 is added to detect black characters from the luminance signal Y and color signals Cr and Cb.
, the binarized luminance signal Y1 and the color signals Crl and Cbl
The present invention is characterized by the addition of a color character area encoding section 52 for encoding, and replacing the character area encoding section 8 of the first embodiment with a black character area encoding section 53. The other components are the same as those in the first embodiment.

The operation of the second embodiment of the present invention configured as described above will be explained below.
, the color space conversion unit 3, the luminance signal binarization unit 4, the color signal generation unit 5, the area determination unit 6, and the halftone encoding unit 7 function in the same manner as in the first embodiment described above, so that the background color detection processing , color space conversion processing, color signal generation processing, and area determination processing operations will be omitted.

The black character determination unit 51 receives the luminance signal Y from the color space conversion unit 3.
and color signals Cr, Cb to predetermined thresholds (Th3, Th4.T
h5) to determine whether each pixel is a black character. That is, if (Y>i' h 3 &CI・<'h 4
&Cb<T h 5 )t h e n k =
l Black text else k=0 Here & represents logical product.

This process is performed for one block, and it is determined whether the block is a black character area or not based on whether the number of black characters is greater than or equal to a predetermined value. That is, 1fK=Σ k>ThG t, h e n black text area else It is determined that the area is other than the black text area.

The determination signal generated by the area determination unit 6 and the black character determination unit 51 is sent to the halftone encoding unit 7, the color character area encoding unit 52, and the black character area encoding unit 53. The halftone encoding unit 7 operates and encodes the luminance signal Y and color signals Crl and Cbl in block jii. As this coding method, for example, the 8r') CT method is used.

On the other hand, if the block is a character area and is a black character area, the black character area encoding @53 operates and the luminance signal 2
The binarized luminance signal Y1 is encoded in units of blocks by the digitizing unit 4. As this encoding method, for example, M l-
Use the 1 force formula.

Further, if it is a character area but not a black character area, that is, if it is a color character area, the color character area encoding unit 52 operates, and the luminance signal Y1 and the color signal Cr1. Cb I is encoded block by block. As for this encoding method, for example, the MH method is used for the luminance signal Y1, and the color signals Crl and Cbl are used.
For example, the ADCT method is used.

The data encoded in this manner is stored or transmitted in the same manner as in the first embodiment. Further, the encoding format is also the same as in the first embodiment. However, there are three types of area information: halftone area information, black text area information, and color text area information.

FIG. 6 is a block configuration diagram of the present invention for decoding -1- encoded data, in which a color character area information section 61 is further added to the configuration of the first embodiment, and The decoding section is replaced with a black character area decoding section 62, and the other parts are the same as those in the embodiment].

For decoding, the encoded data is first subjected to background color information 5 (R1).
, S (G1.), S < Bl) are the base color information decoding section 3
1, and then the area information is decoded by the area information decoding unit 32. For example, if it is a black text area, the black text area decoding unit 62 decodes the black text area (, l] until the next area information is input. According to the following, M+-1 method decoding) is performed. Then, the next area information is read out,
If the information indicates that it is a color text area, the encoded data of the color text area is decoded by the color text area decoding unit 61, and if it is a halftone area, the halftone area decoding unit 34 decodes the coded data of the color text area. Decoding of the encoded data of the area (according to the example described above, AD
CT method decoding).

The decoding process is performed by repeating this process. Then, according to the decoding result of the decoding unit 31, the color signal reproducing unit 35
is the fish signal > 3Cr among the pigs decoded as follows:
Play Cb. However, for the black character area, Cr1
Calculate as =Cbl=O. Furthermore, the luminance signal Y1 used when encoding the character area is used as the luminance signal Y during decoding.

Y=Y] Cr=Crl+0.51S (R1)-0,433(G
l) -0,083(Bl) Cb=Cbl-0,17S (R1)-0,34S(G
]) +0.513 (B l) In general, C=
a R+ b G +c When B, C=CI+
aS (R1) + bS (Gl) + cS (Bl).

The reproduced Cr and Cb and the decoded luminance signal Y are subjected to color space conversion by a color space conversion unit 36 to restore RGB signals.

Note that the encoding format may be such that area information is added to each block, and the block unit is not limited to 8 x 8 pixels, and other encoding methods may be used. May be adopted. Further, the conversion formula from the primary color signal to the luminance signal and color signal is not limited to the embodiment described in 1.

[Effects of the Invention] 1. As explained above, according to the present invention, even if the base of the document is colored, the character part is encoded using a binary encoding method, which is different from the conventional method. An encoding method with a higher compression rate than other encoding methods can be obtained. Furthermore, compared to the conventional method in which encoding is performed as halftones, the edge portions of characters are reproduced more clearly, so the quality of the character portions is significantly improved.

Furthermore, since the color information in the color text area is not lost, the color text area can be reproduced with high quality.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a color image encoding method according to the first embodiment of the present invention, FIG. 2 is a diagram showing the encoding format of the present invention, and FIG. 3 is a diagram for decoding encoded data. FIG. 4 is a block diagram of another embodiment of the present invention provided with a delay buffer, and FIG.
FIG. 6 is a block diagram of a color image encoding system according to a second embodiment of the present invention, and FIG. 6 is a block diagram of a second embodiment of the present invention for decoding encoded data. Binarization unit, 2... base color detection unit, color space conversion unit,
4 - Luminance signal binarization section, color signal generation section, 6...
Area determining unit, halftone encoding unit, 8... character area encoding unit, black character determining unit, - color character area encoding unit, black character area encoding unit, color character area decoding unit, 62 - black character area decoding unit.

Claims (5)

[Claims] (1) In an image encoding method that encodes a color image for each predetermined area, there is a means for detecting the background color of the document, and the value of the color component of the input image signal is corrected according to the detected background color. means for determining whether a predetermined area is a halftone area or a character area based on the corrected color component values; What is claimed is: 1. An image encoding method characterized by comprising means for encoding a luminance component and a color component for a text area, and encoding a luminance component for a character area. (2) In an image decoding method that decodes a color image for each predetermined region, the first means for decoding the underlying color signal from the encoded image data and whether the predetermined region is a halftone region or not. a second means for decoding area information specifying whether it is a character area; a means for selecting a decoding process corresponding to the predetermined area in response to the output of the second means; 2. An image decoding system comprising means for reproducing a color signal in response to the output of the means of item 2. (3) In an image encoding method that encodes a color image for each predetermined area, there is a means for detecting the background color of the document, and the value of the color component of the input image signal is corrected according to the detected background color. a first means for determining whether a predetermined area is a halftone area or a character area from the corrected color component values; and a second means for detecting black characters from the luminance component and the color component. means, and in response to the determination result of the first means,
means for encoding a luminance component and a color component for a halftone region; means for encoding a luminance component for a black character region in response to the output of the second means; 2. An image encoding system characterized by comprising means for encoding binarized luminance components and color components for a color character area in response to the output of the means of item 2. (4) In an image decoding method that decodes a color image for each predetermined region, the first means for decoding the underlying color signal from the encoded image data and whether the predetermined region is a halftone region or not. a second means for decoding area information specifying whether it is a black text area or a color text area; a means for selecting a decoding process corresponding to the predetermined area in response to the output of the second means; An image decoding system comprising means for reproducing a color signal in response to the outputs of the first and second means. (5) An image encoding method for storing or transmitting color image data, characterized in that the color image data is composed of base color information, area information, and encoded data. (6) The image encoding method according to claim (1), characterized in that the background color detection means comprises means for accumulating input images for the time required for detecting the color of the background of the document.