JP2002165105A - Image processing device, its method, and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To raise an image compression ratio by keeping the number of colors small. SOLUTION: When the number of palettes in a character region is 1 (S1901), the character region of the whole binary image is cut out, to generate a partial binary image (S1907). The partial binary image is a subtractive color image in the character region. When the number of the palettes is larger than 1 (S1901), a subtractive color portion 1082 receives an original image 101 and the whole binary image and assigns the value of the pixels in the original image, of which the binary image is black in the character region, to the palette value to make the subtractive color image (S1902). A typical color is extracted for each 1 character, using the subtractive color image and coordinate information for each 1 character, provided with character-dividing information (S1903), and character pixels in the 1 character region are redrawn in the typical color (S1904). This processing is repeated for all the character coordinates, to obtain the final subtractive color image. Above processing is carried out for all the character regions (S1906).

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 and method for compressing an image, and a storage medium.

[0002]

2. Description of the Related Art In recent years, digitalization of documents has been advanced due to the spread of scanners. If the electronic document is held in full color, it becomes about 24 Mbytes in the case of 300 dpi and A4 size, and the memory is too tight to hold, and it is not a size that can be attached to e-mail and sent to others.

On the other hand, JPEG is known for full-color image compression. JPEG is very effective for compressing natural images such as photographs, and has good image quality. However, JPEG compression of high-frequency portions such as character portions causes image degradation called mosquito noise, and the compression ratio is poor. Therefore, the image is first divided into regions, and the JPE
G compression is performed. MMR is used for the character area with color information.
Compression or ZIP compression is performed to obtain an image by each compression.

On the other hand, when these images are decompressed, the white portion is J
There has been a method of transmitting a PEG image and expressing a black portion with a representative character color.

[0005]

However, in the conventional method, the area determined to be a character is subjected to color reduction processing on the entire character area to create a color-reduced image. There is a drawback that a color with a small number of colors is also extracted as a representative color, the number of colors of the reduced color image increases, and the compression ratio deteriorates.

The present invention has been made in view of the above problems, and has as its object to increase the compression ratio of an image by suppressing the number of colors of a color-reduced image.

[0007]

In order to achieve the object of the present invention, for example, an image processing apparatus of the present invention has the following arrangement. In other words, if the image contains a character area,
A character area detecting means for detecting the character area, a base image generating means for generating a base image in which a character part included in the character area is filled with a predetermined color, and a palette creation for creating a color palette representing the character area Means, when the number of color pallets of the character area is one, an image obtained by binarizing the character area is set as a color-reduced image of the character area. In the first color-reduced image of the character area created using the pallet, the representative color of the character portion included in the character area of the first color-reduced image is extracted, and the pixels forming the character part are replaced with the representative color. And a color-reduced image generating means for generating a second color-reduced image of the character area by drawing using the image processing method.

The image processing apparatus further includes a binarizing means for binarizing the image.
The character area detecting means detects a character area included in the image by using a binarized image of the image by the binarizing means.

Further, there is provided a dividing means for dividing the image into regions of a predetermined size.

[0010] The background image generating means may further include an average color calculating means for calculating an average color of an area outside a character portion in the character area, and the character image may be obtained by using the average color by the average color calculating means. Fill the part.

The image processing apparatus further includes a JPEG compression unit for JPEG-compressing the background image by the background image generation unit.

The image processing apparatus further includes a color-reduced image compression means for compressing the color-reduced image of the character area by the color-reduced image generation means by a compression method corresponding to the number of pallets of the color of the character area.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below in detail according to preferred embodiments with reference to the accompanying drawings.

[First Embodiment] FIG. 1 shows a configuration of an image processing apparatus of the present embodiment for creating compressed data 1A, which will be described later,
The processing (processing result) in each section is shown. Reference numeral 101 denotes an original image. An image binarizing unit 102 receives the original image 101 and performs optimal binarization of the image. Reference numeral 103 denotes an entire binary image binarized by the image binarizing unit 102. 1
Reference numeral 04 denotes a character area detection unit that receives the entire binary image 103, detects a character area in the image, and creates character area coordinates 112. Reference numeral 105 denotes a character portion filling portion for extracting an area corresponding to the black area of the entire binary image 103 from the original image 101, filling the area with the surrounding color, and creating an image A. is there. 10
Reference numeral 6 denotes a reduction unit that receives the image A and reduces the size of the image to create an image B. Reference numeral 107 denotes a JPEG compression unit that receives the image B and performs JPEG compression to create a compression code X113. 1
08 inputs the character area coordinates 112, calculates the original image color of the black portion of the entire binary image 103 while referring to the original image 101 and the entire binary image 103 within the coordinates, and sets a plurality of palettes 114. A character color extraction unit that creates and performs color reduction processing of the original image 101 in accordance with the created color. Reference numeral 109 denotes a color-reduced image of a plurality of character areas reduced in color by the character color extraction unit 108. Reference numeral 110 denotes an MMR compression unit that inputs the reduced color image 109 when the reduced color image 109 has 1 bit (m = 1) and performs MMR compression to create a plurality of compression codes Y115. 111 indicates that the color-reduced image 109 is 2 bits or more (m> 1)
Is a ZIP compression unit that inputs the reduced color image 109 and performs ZIP compression to create a plurality of compression codes Z116. Finally, the data from 112 to 116 compiled in 1A are combined to become compressed data.

FIG. 3 shows a flowchart of a process up to the extraction of a character area from the original image 101. Original image 10
1 is a color image.

In step S301, the original image 101 is input, and luminance conversion is performed while thinning the image to reduce the resolution, thereby generating a luminance image J. For example, if the original image is RGB 24-bit, 300 dpi, Y = 0.299R + 0.587G + 0.114B is calculated every four pixels in both the vertical and horizontal directions, and an image J having a luminance value Y is created. Image J
Is an 8-bit 75 dpi image. Step S302
, A histogram of the luminance data of the image J is obtained, and a binarization threshold T is calculated.

In step S303, the luminance image J is binarized by the binarization threshold T to generate a full-scale binarized image 103.
In step S304, contour tracing of black pixels is performed on the entire binary image 103, and labeling is performed on all black regions. In step S305, a character-like area in the black area is determined. This determination is made in Japanese Patent Laid-Open No. 06-068301.
The determination of a character-like area (text block) is performed by a block determination method described in Japanese Patent Application Laid-Open No. H10-260,000.

Next, in step S306, those which are combined from the shape and position of the black area are combined. In this step, the coordinates of each character region and the coordinates of each character region are also detected.

An example of the above processing will be described. For example, when the color original shown in FIG. 4 is input, the histogram of the thinned-out luminance data is obtained as shown in FIG. A threshold T = 1 using data such as average and variance from this histogram.
50 is calculated, and an image obtained by binarizing the image shown in FIG. 4 using the threshold T (the entire image binarized image 103) is as shown in FIG. The contour tracing of the black pixels shown in FIG. 6 is performed, and all of them are labeled. A group of pixels becomes a character area (an image is not actually created, but an image).

If necessary, a group of these black pixels is grouped based on the coincidence of the proximity, the width, and the height.
17 character areas as shown in FIG. 8 can be detected. The coordinate data of these character areas corresponds to the character area coordinates 112 in FIG.
It is.

Next, the original image and the entire binary image (binary image)
An example of the processing in the character part filling unit 105 using is described with reference to each image shown in FIG. 10 and the flowchart of the same processing shown in FIG. FIG. 10A is an original image used below. It is assumed that a binary image of one character area shown in FIG.

In step S1101, the original image is 32 × 3
It divides into two regions (hereinafter, parts) and performs processing for each part. FIG. 10C shows a state where the original image is divided for each part. Since the six parts from 00 to 10 have no character area due to the branch in step S1103, no processing is performed. Since the character area is included in the part No. 11, the process proceeds to step S1104. In step S1104, the average value (average color) ave_color of the RGB values (or YUV may be used) of the color image corresponding to the white portion of this area is calculated with reference to the area of the binary image corresponding to the 11 parts. Step S
At 1105, this ave_color is substituted for the color image corresponding to the black pixel by referring to the corresponding binary image. The above processing is performed by using the character areas 12, 13,
Repeat for parts 21, 22, 23. No processing is performed on other parts since there is no character area. In this way, the average value of the surrounding pixels could be embedded in the portion where the character was present.

This image (image A) is reduced by the reduction unit 106. In the present embodiment, the size is reduced by simple thinning. Incidentally, the order of the reduction process and the character portion filling process may be reversed. In that case, it is necessary to pay attention to the displacement between the binary image and the color image.

Next, FIG. 12 shows a flowchart of the processing in the character color extracting unit 108. The binary image uses the entire binary image 103, but is not limited to this. For example, only the coordinates of the character area and the color image are input, and a representative color calculation process described later is performed using the result of re-binarizing the color image. May be performed.

A counter num representing the number of colors extracted in step S1201 is reset to zero. Since the process is performed for each of the extracted character areas, step S12 is performed.
In 02, it is checked whether or not there is an unprocessed character area coordinate. If there is, the process proceeds to step S1203, and if not, the process ends.

In step S1203, a thinning process is performed on the binary image corresponding to the character area coordinates to reduce black corresponding to a transition from the background to the character portion, thereby creating a new binary image newbi. Next, in step S1204, an RGB three-dimensional histogram of the original image corresponding to the black pixels of the new binary image newbi is obtained. At this time, if a histogram is taken normally, for example, if the input image is 8 bits for each of RGB, a 256 × 256 × 256 histogram is required. It is necessary for the text part to be resolution, no gradation is required, and it is better to ignore some differences in pixel values to calculate the representative color while suppressing variations when reading with a scanner etc. In view of this, such a fine histogram is not necessary. Therefore, in this example, an RGB three-dimensional histogram of the upper 5 bits is taken. When this histogram is taken, the total number of black pixels pixelnum present in the character area is also calculated. In the present embodiment, the RGB space is used.
Other color spaces such as UV may be used. Further, although a three-dimensional histogram is taken, three one-dimensional histograms for each color may be taken.

In step S1205, the maximum value is calculated from the RGB three-dimensional histogram. In this example, the top 5 already
Although a histogram of only bits is taken to suppress noise due to scanner variations, the maximum value of the sum of adjacent values of the histogram is further taken to obtain the original maximum value over two histograms as shown in FIG. Can be detected. More specifically, since the histogram is a three-dimensional histogram, a target point and two histogram values adjacent to each other in the R-axis direction, two adjacent to the G-axis direction, and two adjacent to the B-axis direction are calculated. For example, the maximum value of the total value may be detected. The maximum value detected in this way is represented by colR [num], colG [num], colB
Substitute [num].

In step S1206, step S12
With the maximum value detected at 05 as the center, the histogram value located in the square expanded by, for example, three steps (= haba) is set to 0. The description of the square is illustrated in FIG. FIG. 15 shows a three-dimensional histogram, in which black dots indicate colR [num], colG [num], c
olB [num]. The above-mentioned square is a total of 7 × 7 × 7, which is expanded by three steps around that point. Here, since the three steps are a histogram of the upper 5 bits, they are worth 24 levels with 256 gradations. This is an example, and is not limited to this. After subtracting the histogram value in this square from pixelnum, 0 is substituted. In step S1207, the variable num is incremented by one.

In step S1208, pixelnu
It is checked whether m is equal to or greater than a predetermined threshold thre1, and if m is equal to or greater than thre1, step S120 is performed.
The process proceeds to step S1202 if it is less than 5.

By repeating the above processing for all character areas, a palette for all character areas is created. Next, FIG. 19 shows a flowchart of the processing in the color reduction unit 1082 in the character color extraction unit 108, which will be described below.

If the number of pallets in the character area is 1 (step S1901), the input entire binary image 1
03 is cut out to create a one-part binary image (step S1907). It becomes the color-reduced image of the character area.

On the other hand, if the number of pallets in this area is 2 or more (step S1901), the color reduction unit 1082 inputs the original image 101 and the entire binary image 103, and first uses the character division information without using the character division information. The values of the original image of the pixels where the binary image is black are distributed to the values of the palette, and a color-reduced image is created (step S1902). Then, a representative color is extracted for each character using the reduced color image and the coordinate information for each character based on the character division information (step S1903).
The character pixels in the character area are redrawn in the representative color (step S1904).

Here, the character division information is disclosed in
The method disclosed in Japanese Patent Application Publication No. 068301 discloses a method of obtaining coordinate information for each character by extracting characters. Specifically, line division processing and character cutout processing are performed on a block determined to be a text area to obtain coordinate information (character division information) for each character.

This process is repeated for all the character coordinates, and a color-reduced image 109 is obtained. Then, among the created palettes, palette colors not used in the reduced-color image 109 are deleted from the palette (step S1905). The number of bits to be allocated is 2 bits because 1 data indicating transmission is required as a white portion of the binary image, and when the number of pallets is 3, the transmission data is increased by 1 to 4, which is 2 bits. When the number of pallets is 4, the transmission data is increased by 1 to give 5 bits, which is 3 bits. This number of bits conforms to the number of bits that can be expressed in the finally adapted image format. In this case, for example, if it is determined that the number of pallets is extremely large and the effect of improving the image quality and the compression ratio by multicoloring cannot be seen, the multicoloring may be stopped and the image may be stored as a base image. In that case, the character area must be deleted from the character area coordinate information 112. In addition, before the processing in the character part filling unit 105 is performed, it must be deleted from the character area coordinate information. The processing according to the flowchart shown in FIG. 19 is performed on all the character areas (step S1906).

The color-reduced image 109 thus created
Is 1 bit, the MMR compression unit 110 performs MMR compression to create a compression code Y. If the color-reduced image 109 has two or more bits, the ZIP compression unit 111 performs ZIP compression to create a compression code Z. On the other hand, JPEG compression is performed on the reduced image B to create a compression code X. Character area coordinates 112, palette 114, compression code X11
3. If necessary, a format (compressed data 1A) in which the five compressed codes Y115 and Z116 are combined is created.

If, for example, all the character areas are binary images, the reduced color image created by the character color extraction unit 108 is 1 bit, and in this case, the compression code Z116 is used.
Is not created. That is, the data included in the compressed data 1A is not always limited to the above-mentioned five data, but may be four depending on the character area. This is the same when, for example, all character areas are color images. As an example of a format for combining the five, Adobe PDF or the like can be considered. Adobe PDF is a format that can be displayed by Adobe Reader, an application that Adobe distributes free of charge, and since there is no application that created the document,
Problems such as the inability to open the file on the receiving side can be avoided. Other formats include XML
and so on. XML is a description language for exchanging and distributing documents and data via a network.

FIG. 2 shows the configuration of the image processing apparatus of the present embodiment for expanding the compressed data 1A and the processing in each section thereof.

Reference numeral 201 denotes the input of the compression code X113,
JPEG that performs multi-valued image E by performing PEG decompression processing
Extension means. Reference numeral 202 denotes an enlargement unit that receives the multivalued image E and performs enlargement processing. Reference numeral 203 denotes a multi-value image F enlarged by the enlargement unit 202.

An MMR decompression unit 204 receives the compression code Y115 and creates a binary image G205. 206
Is a ZIP decompression unit that inputs a compression code Z116 and creates a multicolor image H207. 208 is the coordinate 1 of the character area
12 and the corresponding palette 114, and the binary image G205 and / or the multicolor image H207, and when the pixel data of the binary image or the multicolor image indicates transmission, the color of the pixel of the multilevel image F203 is Otherwise, the corresponding palette color is selected and the final image, image I
(Expanded image) This is an image merging unit for creating 209.

FIG. 14 shows the processing steps of the uniting processor 208. First, FIG. 14A shows the result of JPEG expansion of the compression code X. This uses the image of FIG.
When the quantization irreversible method of the G compression is used, the data has pixel values slightly different from those in FIG. But,
Compared with the case where the original image before the character portion is extracted is compressed by the JPEG lossy compression method, the change in the pixel value is small when the same quantization table is used. In this example, the image included in the character area (result of decompression of the compression code Y115) is represented by one bit of a binary image as shown in FIG. 14B, and its palette is R = 20, G = 30, B = 225. Referring to the binary image shown in FIG. 14B, the data of the palette color (20, 30, 255) is placed on the image shown in FIG. An image as shown in FIG. 14 (c) is completed. This becomes the expanded image 209. If the image included in the character area is a multi-color image, the number of palettes changes.
The palettes assigned to the four pixel values 0, 01, 10, and 11 are applied. One of them indicates transmission. For example, if 00 is set, the pixel having the value of 00 is shown in FIG.
4 (a) is selected.

As described above, when the number of pallets in a character area is one, the character area is represented using a binary image by the image processing apparatus and method according to the present embodiment. On the other hand, when the number of pallets is two or more, each character is represented by each representative color. As a result, a color with a small ratio used in the character area is not included in the color-reduced image, so that an image with a small number of colors and high compression efficiency can be generated.

[Second Embodiment] In the first embodiment, binarization of an original image (color image) is performed as a character area extraction process. However, the present invention is not limited to this. Calculate the edge amount of the pixel with neighboring pixels,
There is a method in which a binary image obtained by binarizing the edge amount is similarly subjected to contour tracing to detect a character region. The configuration of an image processing apparatus that executes this method is as shown in FIG.

In the example of the configuration shown in FIG.
No. 3 cannot be used for filling a character portion and extracting a character color, so a binary image (partial binary image 1706) is used for each character region.
Create This binary image is stored in, for example, the character area detecting section 1.
The binarization may be performed using the threshold value T calculated in step 704, or a histogram may be obtained for each area to calculate an optimum binarization threshold value for the character area. Compared with the histogram of the entire surface shown in FIG. 5, the luminance histogram of a part of the character area can be expected to have a simple shape as shown in FIG. 9, so that the threshold can be easily determined. Reference numeral 901 denotes a set of background colors, and reference numeral 902 denotes a set of character colors.

In the first embodiment, it is impossible to process a high-luminance character (reversed character) on a low-luminance background. However, if a character area is detected by this differential processing, the reversal character area can also be detected. It becomes possible. 1702 is a differential processing unit,
A differential filter as shown in FIG. 17 is applied to the center of the pixel of interest, and if the absolute value exceeds the threshold value, it is binarized, such as black, and if not, white. FIG. 17A shows a primary differential filter. The upper part can detect a horizontal line, and the lower part can detect a vertical line. The diagonal line can be detected by using the sum of the absolute values of the two filters. Further, an oblique line filter may be used. FIG. 17 (b)
Is a second derivative filter and corresponds to all directions. The second derivative filter can also be created in the horizontal direction and the vertical direction. Such a filter is applied to all pixels to create a differential binarized image 1703. At this time, it is also possible to simultaneously lower the resolution by applying a filter while thinning out the pixels. If the processing from step S304 in FIG. 3 is performed on the binary image created as described above, the character area coordinates including the inverted character can be detected.

In the configuration that also targets the inverted character, the binarizing unit 1705 must also correspond to the inverted character. First
In the embodiment, it is assumed that only the pattern shown in FIG. 9 is included. However, in the case of this configuration in which the inverted character area can be extracted as the character area, the three patterns shown in FIG. Become. FIG. 18B is a diagram showing a luminance distribution of a character and a background in the case of a reversed character. FIG.
FIG. 3C is a diagram illustrating respective luminance distributions when two colors of black characters and white characters exist on the same gray base. Considering these three patterns, the binarization unit 1705 detects the points A and B, and preferably performs the binarization processing of white between the regions A and B and black for the other regions. Alternatively, without considering the pattern shown in FIG. 18C, one threshold value for separating the background and the character portion may be detected, and if the pattern is an inversion pattern, the inversion processing may be performed.

As described above, if the reversed character area also corresponds,
The inverted character area remaining in the first embodiment on the EG-compressed image is also smoothed by filling the character portion, so that the compression efficiency is good, and the inverted character portion also has no degradation in resolution or mosquito noise. Can be compressed.

In the first embodiment and the present embodiment using the above-described differentiation, the character region extraction processing is performed using a binary image. However, the present invention is not limited to this. The pixel value of the multi-value image itself is referred to. The character area may be guessed.

In the first embodiment, the degree of reduction in the reduction unit 106 is constant for all images. However, the present invention is not limited to this. For example, a reduction parameter determination unit that determines reduction parameters (for example, 、 ２, 、 ４) of resolution conversion may be provided. As a method of realizing this, for example, orthogonal transformation is performed on the entire surface of the image A every 8 × 8,
If there is an area where the coefficient of the high-frequency part of the orthogonal transformation result is larger than the threshold value, the reduction can be adjusted to 、, and if the area is smaller than the threshold, the reduction can be adjusted to ４. This parameter is not limited to two steps, but may be, for example, three steps (one-half, one-fourth without reduction). This has the effect of avoiding extreme reduction of the high frequency part and preventing image quality degradation. In order to determine the reduction parameter, a method of applying a differential filter to the image and switching from the sum of the absolute values is also conceivable. For example, if the sum of the differences between adjacent pixel values is equal to or greater than a threshold value m, no reduction is performed, if n is equal to or greater than n, 縮小, if less than n, ４.

The configuration of the apparatus for decoding the compressed data 1A by the image processing apparatus in the present embodiment may be an apparatus having the configuration shown in FIG.

[Other Embodiments] In the embodiment described above, the number of tiles does not need to be plural as shown in FIG. 2 or FIG. 5, and the processing may be performed with the entire image as one tile. In this case, it is not necessary to extract a feature amount for each tile.

Further, the present invention is not limited to only the apparatus and method for realizing the above-described embodiment, but includes a computer (CPU or MPU) in the above-mentioned system or apparatus.
The present invention is also applicable to the case where the program code of software for realizing the above-described embodiment is supplied, and the computer of the system or the apparatus operates the above-described various devices according to the program code to realize the above-described embodiment. include.

In this case, the program code of the software implements the functions of the above-described embodiment, and the program code itself and means for supplying the program code to the computer, specifically, A storage medium storing the program code is included in the scope of the present invention.

Examples of a storage medium for storing such a program code include a floppy (registered trademark) disk, a hard disk, an optical disk, a magneto-optical disk, and a CD.
-A ROM, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used.

In addition to the case where the computer controls the various devices according to only the supplied program code to realize the functions of the above-described embodiment, not only the case where the program code runs on the OS Such a program code is also included in the scope of the present invention when the above embodiment is realized in cooperation with an (operating system) or other application software.

Further, after the supplied program code is stored in a memory provided in a function expansion board of a computer or a function expansion unit connected to the computer, the function expansion board or the function is stored based on the instruction of the program code. The case where the CPU or the like provided in the storage unit performs part or all of the actual processing, and the processing implements the above-described embodiment is also included in the scope of the present invention.

When the present invention is applied to the above-mentioned storage medium, the storage medium has the above-described structure (FIG. 3, and / or FIG. 1).
1 and / or FIG. 12 and / or FIG. 19).

[0057]

As described above, according to the present invention, the compression ratio of an image can be increased by suppressing the number of colors of a color-reduced image.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an image processing apparatus that creates compressed data 1A, which will be described later, according to a first embodiment of the present invention, and a process (process result) in each unit thereof.

FIG. 2 shows compressed data 1 according to the first embodiment of the present invention.
FIG. 2 is a block diagram illustrating a configuration of an image processing apparatus that expands A and a process in each unit thereof.

FIG. 3 is a flowchart of a process until a character area is extracted from an original image 101;

FIG. 4 is a diagram illustrating a color document used in the first embodiment of the present invention.

FIG. 5 is a histogram of luminance values of the color original shown in FIG. 4;

FIG. 6 is a diagram showing an image obtained by binarizing the color original shown in FIG. 4;

7 performs contour tracing of black pixels in the image shown in FIG. 6,
FIG. 11 is a diagram illustrating an image in which all are labeled and only a group of black pixels whose horizontal width is equal to or less than a predetermined threshold or whose height is equal to or less than a predetermined threshold is allowed as a character.

8 is a diagram showing an image in the case where a group of black pixels is grouped based on the closeness of position, the width, and the coincidence of height in the image shown in FIG. 7;

FIG. 9 is a histogram for explaining luminance values of a base color and a character color in the second embodiment of the present invention.

10A is a diagram illustrating an original image, FIG. 10B is a diagram illustrating a binary image of a character area, and FIG. 10C is a diagram illustrating a state in which the original image illustrated in FIG.

FIG. 11 is a flowchart of a process performed by a character part painting unit 105 using an original image and an entire binary image (binary image).

FIG. 12 is a flowchart of a process in a character color extraction unit 108;

FIG. 13 is a diagram showing an original maximum value over two histograms.

FIG. 14A shows a result of JPEG expansion of a compression code X,
(B) is the result of decompression of the compression code Y (binary image), (c)
FIG. 4 is a diagram showing an image in which data of the image shown in FIG.

FIG. 15 is a diagram for explaining a square for detecting a maximum value in an RGB three-dimensional histogram.

FIG. 16 is a block diagram illustrating a configuration of an image processing apparatus according to a second embodiment of the present invention.

17A is a diagram illustrating a primary differential filter for detecting vertical and horizontal lines, and FIG. 17B is a diagram illustrating a secondary differential filter corresponding to all directions.

FIGS. 18A to 18C are diagrams showing luminance distributions of three patterns of an image of a character area and a base image.

FIG. 19 is a flowchart of processing in a color reduction unit 1082 in the character color extraction unit 108.

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) H04N 1/46 Z F Term (Reference) 5B057 CA01 CA08 CA12 CA16 CB01 CB06 CB08 CB12 CB18 CE06 CE12 CE17 CG01 DB02 DB09 DC17 DC23 5C077 LL20 MP05 MP06 MP08 PP21 PP27 PP28 PP31 PP46 PP47 PP49 PQ12 PQ18 RR02 RR06 RR21 5C078 AA09 BA27 BA57 CA02 CA03 DB04 DB05 DB06 5C079 LA02 LA06 LA14 LA26 LA34 LA39 LB12 MA11 NA00

Claims (30)

[Claims] 1. A character area detecting means for detecting a character area when an image includes a character area, and a base image generating means for generating a base image in which a character part included in the character area is filled with a predetermined color. Means, a palette creating means for creating a palette of colors representing the character area, and when the number of color palettes in the character area is one, a binarized image of the character area is set as a reduced color image of the character area, When the number of color palettes of the area is plural, in the first color-reduced image of the character area created using the color palette of the character area, a representative of the character portion included in the character area of the first reduced color image An image processing apparatus comprising: a color-reduction image generating unit that generates a second color-reduced image of the character area by extracting a color and drawing a pixel configuring the character portion using the representative color. . 2. The image processing apparatus according to claim 1, further comprising a binarizing unit for binarizing the image, wherein the character region detecting unit detects a character region included in the image using the binarized image of the image by the binarizing unit. The image processing apparatus according to claim 1, wherein: 3. The image processing apparatus according to claim 1, wherein the character area detection unit detects a character area included in the image from an image obtained by performing a differentiation process on the image. 4. The character area detecting means calculates an edge amount of all pixels with neighboring pixels by applying a differential filter to the image, and performs contour tracing on an image obtained by binarizing the edge amount. The image processing apparatus according to claim 3, wherein a character area is detected by performing the detection. 5. The image processing apparatus according to claim 1, further comprising a division unit configured to divide the image into regions each having a predetermined size. 6. The base image generating means further includes an average color calculating means for calculating an average color of an area outside a character part in a character area, and the character image is formed by using the average color by the average color calculating means. The image processing apparatus according to claim 1, wherein the section is filled. 7. The apparatus according to claim 1, further comprising JPEG compression means for JPEG-compressing the background image by said background image generation means. 8. The image processing apparatus according to claim 7, wherein the background image compressed by said JPEG compression means is reduced. 9. The base image compressed by the JPEG compression unit is reduced at a reduction rate corresponding to a region where a coefficient of a high-frequency part is large when the base image is orthogonally transformed. An image processing apparatus according to claim 1. 10. The image processing apparatus according to claim 1, further comprising a color-reduced image compressing means for compressing the color-reduced image of the character area by the color-reduced image generating means by a compression method according to the number of color palettes of the character area. The image processing apparatus according to any one of the preceding claims. 11. The color-reduced image generating means performs MMR compression on the color-reduced image of the character area when the number of color pallets in the character area is one, and when the color pallet number of the character area is plural, ZIP for the color-reduced image of the character area
The image processing apparatus according to claim 10, wherein compression is performed. 12. The image processing apparatus according to claim 1, wherein a palette of a color not used by the color-reduced image generating means is deleted from the palettes by the palette creating means. 13. When an image includes a character area,
A character area detecting step of detecting the character area, a base image generating step of generating a base image in which a character part included in the character area is filled with a predetermined color, and a palette creation of creating a color palette representing the character area If the number of color palettes of the character area is one, the binarized image of the character area is used as the color-reduced image of the character area. In the first color-reduced image of the character area created using the pallet, the representative color of the character portion included in the character area of the first color-reduced image is extracted, and the pixels forming the character part are replaced with the representative color. A color-reduced image generating step of generating a second color-reduced image of the character area by drawing using the image processing method. 14. The image processing apparatus further comprising a binarization step of binarizing the image, wherein the character area detection step detects a character area included in the image using a binarized image of the image obtained by the binarization step. 14. The image processing method according to claim 13, wherein: 15. The image processing method according to claim 13, wherein in the character region detecting step, a character region included in the image is detected from an image obtained by performing a differentiation process on the image. 16. The image processing method according to claim 13, further comprising a dividing step of dividing the image into regions each having a predetermined size. 17. The base image generation step further includes an average color calculation step of calculating an average color of an area outside a character part in a character area, and using the average color in the average color calculation step to generate the character. 14. The image processing method according to claim 13, wherein a part is filled. 18. The image processing method according to claim 13, further comprising a JPEG compression step of JPEG-compressing the background image in the background image generation step. 19. The image processing apparatus according to claim 13, further comprising a color-reduced image compression step of compressing the color-reduced image of the character area in the color-reduced image generation step by a compression method according to the number of color palettes of the character area. The image processing method described in the above. 20. In the reduced color image generation step, when the number of color pallets in the character area is one, MMR compression is performed on the reduced color image in the character area, and when the number of color pallets in the character area is plural, ZI for the color-reduced image of the character area
The image processing method according to claim 19, wherein P compression is performed. 21. The color palette not used in the reduced color image generation step is deleted from the palettes in the palette generation step.
The image processing method according to 1. 22. A computer-readable storage medium storing a program code, wherein, when a character area is included in an image, a program code of a character area detecting step of detecting the character area; A program code for a background image generating step of generating a base image in which a character portion included in the character area is filled with a predetermined color; a program code of a palette creating step for generating a palette of colors representing the character area; When the number of pallets is one, an image obtained by binarizing the character area is used as a color-reduced image of the character area. When the number of pallets of the color of the character area is plural, the image created using the palette of colors of the character area is used. In the first color-reduced image of the character region, the representative color of the character portion included in the character region of the first color-reduced image is extracted, and the pixels forming the character portion are extracted. Storage medium, comprising a program code of reduced color image generation step of generating a second color-reduced image of the character region by drawing using a color. 23. The image processing apparatus further comprising a program code for a binarizing step of binarizing the image, wherein the character area detecting step uses a binarized image of the image obtained by the binarizing step in a character area included in the image. 23. The storage medium according to claim 22, wherein the storage medium is detected. 24. The storage medium according to claim 22, wherein in the character area detecting step, a character area included in the image is detected from an image obtained by differentiating the image. 25. The storage medium according to claim 22, further comprising a program code for a dividing step of dividing the image into regions of a predetermined size. 26. The program code of the background image generating step further includes a program code of an average color calculating step of calculating an average color of an area outside a character part in a character area, 23. The storage medium according to claim 22, wherein the character portion is filled using a color. 27. The storage medium according to claim 22, further comprising a program code of a JPEG compression step of JPEG-compressing the base image in said base image generation step. 28. The program according to claim 28, further comprising a program code for a color-reduced image compression step of compressing the color-reduced image of the character area in the color-reduced image generation step by a compression method according to the number of color palettes of the character area. Item 23. The storage medium according to Item 22. 29. In the reduced color image generation step, when the number of color pallets in the character area is one, MMR compression is performed on the reduced color image in the character area, and when the number of color pallets in the character area is plural, ZI for the color-reduced image of the character area
The storage medium according to claim 28, wherein P compression is performed. 30. A pallet of a color not used in the subtractive color image generation step is deleted from pallets in the pallet creation step.
A storage medium according to claim 1.