JP4693289B2 - Image compression apparatus, image compression method, program code, and storage medium - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image compression apparatus, an image compression method, a program code, and a storage medium for compressing a color document image.
[0002]
[Prior art]
In recent years, the digitization of documents has progressed with the spread of scanners. If an electronic document is owned in full color, it will be about 24 Mbytes in 300 dpi and A4 size, and it will not be a size that can be sent to others by attaching a mail, etc. 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, when JPEG compression is performed on a high-frequency portion such as a character portion, image degradation called mosquito noise occurs, and the compression rate is also poor. Therefore, the original image is divided into regions to create a JPEG compressed image of the background portion from which the character region is removed and an MMR compressed image of the character region portion with color information.
[0003]
However, in the above method, for example, an image including a character part using two or more colors is compressed by the above compression method, such as information on highlighted characters indicated by red in a black character sentence is lost, and the compressed image Is expanded, the character portion included in the expanded image is set to one color.
[0004]
On the other hand, as a method for highly compressing a color document image with little deterioration in image quality, there has been the following method. First, a binarization process is performed on a color document image to obtain a binary image. A character area is detected from the binary image. Specifically, the black pixel contour line in the binary image is traced, and all black regions are labeled. Then, the labeled black area is searched to determine an area that seems to be a character in the black area. Then, an area in the original image corresponding to the black area of the binary image is filled with a color around the black area to create an image A. Then, the image B obtained by reducing the image A is JPEG compressed. Then, the color of the area of the original image (color document image) corresponding to the black area of the binary image is calculated to create a plurality of palettes. Further, a color reduction process is performed on the original image according to the palette to generate a color reduction image. When the reduced color image is 1 bit, the reduced color image is subjected to MMR compression. When the reduced color image is 2 bits or more, the reduced color image is reversibly compressed.
[Problems to be solved by the invention]
However, in the conventional method, when there are many text parts in the image, the partial compression size of the text area is small, but header information such as position coordinates and color information (palette) of the partial image is added to each. As a result, there is a drawback that the compression size increases. For example, in the case of an image in which a table is arranged on the entire surface, each cell of the table is processed as a text area, so the file size after compression becomes large.
[0005]
The present invention has been made in view of the above problems, and an object of the present invention is to suppress the size of compressed data obtained by compressing a color document image including a text area.
[0006]
[Means for Solving the Problems]
In order to achieve the object of the present invention, for example, an image compression apparatus of the present invention comprises the following arrangement.
[0007]
That is, an image compression apparatus for compressing a color document image,
Extracting means for extracting the color of the text area included in the color document image;
Generating means for generating an inclusion image including a text area having a color within a preset color range in the text area;
Compression means for compressing the inclusion image and / or the text region ,
The generating unit includes a determining unit that determines a text region to be combined with a target text region among text regions having a color within a preset color range,
Generating an inclusion image including the attention text area and the text area determined by the determination unit; and information about the inclusion image;
The determining means includes
The first compressed size obtained by the sum of the compressed size estimated when the target text area or the inclusion image including the target text area is compressed and the compressed size estimated when the text area is compressed When,
Obtaining a second compressed size estimated when the attention image area or an inclusion image including the attention text area and an inclusion image including the text area are compressed;
When the second compressed size is smaller than the first compressed size, the text region is combined with the target text region .
[0011]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, the present invention will be described in detail according to preferred embodiments with reference to the accompanying drawings.
[0012]
[First Embodiment]
FIG. 1 shows a basic configuration of an image compression apparatus according to this embodiment. A CPU 101 controls the entire apparatus using programs and data stored in the RAM 102 and the ROM 103 and performs image compression processing described later. Reference numeral 102 denotes a RAM which has an area for temporarily storing programs and data loaded from the external storage device 104 and the storage medium drive 109, image data input from the image input device 108, and the like, and the CPU 101 performs various processes. It also has a work area used for execution. A ROM 103 stores a control program and a boot program for the entire apparatus, setting data for the apparatus, and the like. Reference numeral 104 denotes an external storage device such as a hard disk, which can store programs and data loaded from the storage medium drive 109. When the size of the work area exceeds the size of the RAM 102, the excess area can be provided as a file. Reference numerals 105 and 106 denote a keyboard and a mouse, respectively, which function as pointing devices, and can input various instructions to the apparatus.
[0013]
Reference numeral 107 denotes a display device, which is composed of a CRT, a liquid crystal screen, or the like, and can display image information and character information. An image input device 108 includes a scanner, a digital camera, and the like, and can input an image as data. Note that the image input apparatus 108 includes an interface for connecting to the apparatus. Reference numeral 109 denotes a storage medium drive, which includes a CD-ROM drive, DVD-ROM drive, floppy disk (FD) drive, and the like, and reads programs and data from a storage medium such as a CD-ROM, DVD-ROM, or FD. be able to. A bus 110 connects the above-described units.
[0014]
FIG. 2 shows a functional configuration of the image compression apparatus according to this embodiment. Reference numeral 201 denotes a color document image, and a plurality of colors are used for a character portion (text portion) included in the image 201. A binarization unit 202 binarizes the color document image 201 to generate a binary image. An area analysis unit 203 identifies a text area in the binary image and generates information (text information) such as the position and size of the text area in the binary image. As a method for specifying a text area, for example, the outline of a black pixel in a binary image is traced, all the black areas are labeled, the labeled black area is searched, and characters in the black area are likely to be identified. A method for specifying a region (that is, a region that seems to be a text region) can be mentioned, but the method is not limited to this.
[0015]
A text part color extracting unit 204 extracts a color used for each text area. An image combining unit 205 generates an image of an area including a text area determined to use the same color (hereinafter referred to as an included image). A binary image compressing unit 206 compresses the included image and / or the text area generated by the image combining unit 205. Reference numeral 207 denotes a character portion filling unit that generates an image (hereinafter referred to as a background image) in which the text region analyzed by the region analysis unit 203 is filled with a predetermined color in the color document image 201. The predetermined color may be a predetermined color or an average value of pixels around the text area. A background image compression unit 208 compresses the background image generated by the character portion painting unit 207.
[0016]
2 is stored in a storage medium, and this storage medium is stored in the image compression apparatus shown in FIG. 1 via the storage medium drive 109 (via the storage medium drive 108). It may be read. In this case, when the CPU 101 executes the read program, the image compression apparatus having the configuration shown in FIG. 1 operates as an apparatus having the functional configuration shown in FIG.
[0017]
A color document image compression method according to this embodiment will be described with reference to the functional configuration diagram of FIG.
[0018]
First, the color document image 201 is read into the RAM 102 from any of the external storage device 104, the image input device 108, or the storage medium drive 109. In the present embodiment, an image shown in FIG. 3A is used as the color document image 201.
[0019]
Next, based on the color document image 201 read into the RAM 102, the binarization unit 202 generates a binary image. A method for generating a binary image is not particularly limited, but in the present embodiment, the following method is used. First, a histogram of luminance data in the color document image 201 is taken, and a binarization threshold T is calculated. Although this calculation method is not particularly limited here, for example, a luminance value that is an intermediate value of the histogram may be used as the threshold value T. Then, the color document image 201 is binarized using the binarization threshold T, and a binary image is created. The generated binary image is stored in a different area from the area where the color document image 201 is stored in the RAM 102.
[0020]
Next, the region analysis unit 203 specifies the text region by referring to the above binary image by the above method. At that time, the above-described text information is generated. When the object whose area is analyzed by the area analysis unit 203 is the image shown in FIG. 3A, the result is shown in FIG. 3B. In the figure, TEXT1 to TEXT5 are specified as text areas, and text information is generated for each area. This text information is output to the text part color extracting part 204 and the character part filling part 207.
[0021]
The text part color extracting unit 204 refers to the text information, identifies the text area in the color document image 201, and extracts the color in the identified text area, that is, the color of the character in the text area. 3B, in this embodiment, the TEXT1 and TEXT3 areas are red characters, the TEXT2 and TEXT4 areas are black characters, and the TEXT5 is blue characters. The color of each text area extracted by the text part color extracting unit 204 is generated as palette information.
[0022]
The image combining unit 205 combines the text regions using the same color with reference to the palette information described above for the colors in the text regions TEXT1 to TEXT5 extracted by the text portion color extracting unit 204. In this case, since TEXT 1 and TEXT 3 include characters using the same color, an image of an area including these areas (an included image) is generated. Here, generating an image of an area including TEXT1 and TEXT3 (including image) is referred to as “joining TEXT1 and TEXT3”. This inclusion image is shown as TEXT 1 ′ in FIG. 3C. It should be noted that the pixels in the included image have monochrome pixel values except for the character portion. The same applies to TEXT2 and TEXT4. An inclusion image including TEXT2 and TEXT4 is indicated by TEXT2 ′ in FIG. 3C. Details of the inclusion images TEXT1 ′ and TEXT2 ′ are shown in FIG. 3D, respectively. The image combining unit 205 also generates inclusion image information including the position and size (in the binary image or the color document image 201) of each inclusion image.
[0023]
A method for specifying a text region using the same color in the image combining unit 205 will be described. When the text color in the text area is 8 bits for each RGB, the color is reduced to a predetermined color range such as 2 bits for each RGB or 3 bits for each RGB. Each text area is reduced in this way, and it is determined whether or not the same color is obtained. The degree of color reduction depends on how much gradation the compressed image is desired to have. For example, RGB may be 2 bits, 2 bits, and 1 bit, respectively, or RGB may be 3 bits, 3 bits, and 2 bits by utilizing the low sensitivity to the blue color of the human eye.
[0024]
If it is desired to determine the same color more accurately, it is preferable to convert the color difference to the LAB format or the YCrCb format for easier comparison of the color difference and round it to 2 bits or 3 bits. To explain, in the RGB format, when comparing black with gray and dark blue, the dark blue becomes closer in distance, but in the LAB and YCrCb formats, the luminance component and the color component are separated, so the black and dark blue can be separated. It becomes possible.
[0025]
The color of the scanned character is slightly different, but if the color is low, such as black characters, the color with the lowest luminance in the text area of the same color is used, and conversely, if the color is high, such as white characters, etc. If the brightest color in the text area of the same color is adopted, the reproducibility of the input image is somewhat lowered but the appearance is improved.
[0026]
The binary image compression unit 206 compresses each included image and / or text region, but there may be a text region having a plurality of colors. Therefore, when compressing a text area, the compression method is changed depending on whether the text area has one color or a plurality of colors. This is determined by referring to the palette information in the text area. As a result of referring to the palette information, if the target text area has only one color, MMR compression is performed on the target text area. If the target text area has a plurality of colors, the target text area Reversible compression is performed. Further, the above palette information and text information are attached to the compression result as a header.
[0027]
On the other hand, when compressing an included image, MMR compression is used. In addition, the palette information and the included image information of the included image are attached to the compression result as a header. Note that palette information exists for each text area, but all text areas in the included image have the same palette information. Therefore, any one palette information of the text area in the inclusion image may be used as the palette information of the inclusion image.
[0028]
In this way, when each text area is compressed, five headers (headers for TEXT1 to TEXT5) are created, whereas in this embodiment, three headers (headers for TEXT1 ′, TEXT2 ′, and TEXT5) are created. Will be created. As a result, the number of headers can be reduced, and as a result, the size of the compressed data is reduced.
[0029]
On the other hand, the character portion filling unit 207 identifies a text region in the color document image 201 using text information, and generates an image (background image) in which the identified text region is painted with a predetermined color. This base image is shown in FIG. 3E. The predetermined color may be a predetermined color or may be an average value of pixels around the text area in the color document image 201.
[0030]
Then, the background image compression unit 208 performs JPEG compression on the image (background image) generated by the character portion painting unit 207.
[0031]
As described above, even when a color document image including many text areas is compressed by the image compression apparatus and the image compression method according to the present embodiment, an image including a text area having the same color is generated, Since compression is performed, the number of headers attached to the compressed image can be reduced. At the same time, the size of the compressed data can be reduced.
[0032]
[Second Embodiment]
In the first embodiment, text regions having the same color are included in the same inclusion image and subjected to MMR compression. However, when a small text area that is the same color but is included in the included image is included, the size after compression may increase. In this embodiment, an image compression method in such a case will be described below.
[0033]
The functional configuration of the image compression apparatus in the present embodiment is different from the first embodiment in the processing in the image combining unit 205 in the functional configuration diagram shown in FIG. Therefore, the processing of the image combining unit 205 in the present embodiment will be described with reference to FIG.
[0034]
FIG. 4 is a flowchart of specific processing in the image combining unit 205 of the present embodiment.
[0035]
First, the image combining unit 205 selects one reference text area (hereinafter referred to as a reference text area) from the text area group determined to have the same color (step S401). If there is no text area, or if the process described later has been completed for all text areas (step S402), this process ends. On the other hand, if there is an unprocessed text area, the process proceeds to step S403.
[0036]
A text area in the vicinity of the reference text area that has the same color is searched (step S403). If there is a suitable area that matches this condition, the process proceeds to step S404, and the text that matches this condition is found. A text area closest to the reference text area (hereinafter referred to as a neighboring text area) is selected (step S404). On the other hand, if there is no text area that matches the above-described conditions, the process proceeds to step S409, and an inclusion image including the text area considered to be combined with the reference text area in step S408 described later is created (step S408). S409).
[0037]
Next, an inclusion image rectangle including the reference text region and the neighboring text region is determined (step S405). Then, when each of the reference text area and the neighboring text area is compressed, the total size of the respective compressed data and the compressed size when the inclusion image is compressed are estimated (step S406). Here, there is a method of actually compressing and obtaining an accurate size, but if it is simply calculated by the following method, the processing time can be reduced although the accuracy of the compression size is reduced. When the two areas (reference text area and neighboring text area) are compressed using the compression ratio A of the text area measured in advance, the total size can be estimated by the following equation.
[0038]
Compression size 1 = (area of reference text area + area of adjacent text area) × A + 2 × header size On the other hand, when compressing an included image, two areas included in the included image, the reference text area and the adjacent text area There is always a gap. This portion is filled with data representing a single pixel value, and can be compressed at a much higher compression rate than when the text area is compressed. If this compression ratio is B, the compression size is 2 = (area of text area) × A + (area of gap) × B + header size.
[0039]
Then, the compression size 1 and the compression size 2 are compared using the above estimation result, and the compression size 2 is smaller, that is, the case where the inclusion image is compressed is generated than the case where each region is compressed separately. If the size of the compressed data to be reduced is small (step S407), the process proceeds to step S408, and data indicating that the reference text region and the neighboring text region are included (combined) in the same inclusion image in the combined list (added) (step S407). Step S408).
[0040]
FIG. 5 shows an example of a combined list. In the figure, a configuration example of the combined list when the reference text area is TEXT2 is shown, and correspondence between TEXT2 and each text area TEXT1 to TEXT5 is shown. In the figure, 0 is a code indicating that they are not combined, 1 is a code indicating that they are combined, and 999 is a code indicating invalid (cannot be combined with itself). A code (0 in the figure) indicating that all items are not initially combined is set in the combination list, and a code (1 in the figure) indicating that the items are combined only when the process in step S408 is executed. Changed to
[0041]
On the other hand, if the compressed size 2 is larger, that is, if the size of the generated compressed data is larger when compressing the included image than when compressing each area separately (step S407), the process goes to step S403. Go back and search the next neighborhood text area.
[0042]
When the above process is completed and the reference text area and the adjacent text area are combined, the process after step S403 is performed again, except for the text area selected once, and the same color as the reference text area. The text area closest to the reference text area is set as a new neighboring text area (steps S403 and S404). Then, an inclusion image (second inclusion image) rectangle including the reference text area, the previous neighborhood text area, and the current neighborhood text area is determined (step S405), and the second inclusion image and the current neighborhood text area are determined. , The compression size 1 and the compression size 2 are estimated using the above formula (step S406). Specifically, the following equation is obtained.
[0043]
Compression size 1 = (Area of second inclusion image + Area of neighboring text area) × A + 2 × Header size Compression size 2 = (Area of text area) × A + (Area of gap) × B + Header size The process after step S407 is performed. By doing so, it is possible to create an inclusion image that includes the most text regions and has the smallest size after compression.
[0044]
[Other Embodiments]
Further, the present invention is not limited to only the apparatus and method for realizing the above-described embodiment, and a software program for realizing the above-described embodiment on a computer (CPU or MPU) in the system or apparatus. A case where the embodiment is realized by supplying a code and causing the computer of the system or apparatus to operate the various devices according to the program code is also included in the scope of the present invention.
[0045]
In this case, the software program code itself realizes the functions of the above embodiment, and the program code itself and means for supplying the program code to the computer, specifically, the program code is stored. The storage medium is included in the category of the present invention.
[0046]
As a storage medium for storing such a program code, for example, a floppy disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used.
[0047]
The computer controls various devices according to only the supplied program code, so that not only the functions of the above embodiments are realized, but also the OS (operating system) on which the program code is running on the computer. In the case where the above embodiment is realized in cooperation with other application software or the like, such program code is also included in the scope of the present invention.
[0048]
Further, after the supplied program code is stored in the memory of the function expansion board of the computer or the function expansion unit connected to the computer, the program code is stored in the function expansion board or function storage unit based on the instruction of the program code. A case in which the CPU or the like provided performs part or all of the actual processing and the above-described embodiment is realized by the processing is also included in the scope of the present invention.
[0049]
【The invention's effect】
As described above, according to the present invention, the number of headers provided for each text area is generated by generating an inclusion image including a text area having the same color within a predetermined color range and a header of the inclusion image. The number of headers of the included image can be reduced. As a result, the size of the compressed data obtained by compressing the color document image including the text area can be suppressed.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating a basic configuration of an image compression apparatus according to a first embodiment of the present invention.
FIG. 2 is a diagram illustrating a functional configuration of the image compression apparatus according to the first embodiment of the present invention.
3A is a diagram showing a color document image 201. FIG.
3B is a diagram showing a text area of the color document image 201 specified by the area analysis unit 203. FIG.
FIG. 3C is a diagram showing an inclusion image.
FIG. 3D is a diagram showing details of TEXT1 ′ and TEXT2 ′.
FIG. 3E is a diagram showing a base image.
FIG. 4 is a flowchart of specific processing in an image combining unit 205 according to the second embodiment of this invention.
FIG. 5 is a diagram illustrating an example of a combined list.

Claims (16)

An image compression apparatus for compressing a color document image,
Extracting means for extracting the color of the text area included in the color document image;
Generating means for generating an inclusion image including a text area having a color within a preset color range in the text area;
Compression means for compressing the inclusion image and / or the text region ,
The generating unit includes a determining unit that determines a text region to be combined with a target text region among text regions having a color within a preset color range,
Generating an inclusion image including the attention text area and the text area determined by the determination unit; and information about the inclusion image;
The determining means includes
The first compressed size obtained by the sum of the compressed size estimated when the target text area or the inclusion image including the target text area is compressed and the compressed size estimated when the text area is compressed When,
Obtaining a second compressed size estimated when the attention image area or an inclusion image including the attention text area and an inclusion image including the text area are compressed;
An image compression apparatus for combining the text area with the target text area when the second compression size is smaller than the first compression size . Further, the image processing apparatus includes binarization means for performing binarization processing on the color document image and generating a binary image,
The image according to claim 1, wherein the extraction unit specifies a text region from the binary image, and extracts a color included in a region corresponding to the specified text region in the color document image. Compression device.   The image compression apparatus according to claim 2, wherein the extraction unit generates text information including a position of a text area and a size of the text area in the binary image.   The image compression apparatus according to claim 2, wherein the extraction unit generates a color extracted from the text area as palette information of the text area.   The generating means performs a color reduction process on the colors in the text area, and includes an included image including a text area having a color within a preset color range in the color subjected to the color reduction process, and information on the included image The image compression apparatus according to claim 1, wherein the image compression apparatus is generated.   6. The image compression apparatus according to claim 5, wherein the information on the inclusion image includes a position of the inclusion image in the color document image and a size of the inclusion image. The generating means generates a combined list indicating a text area to be combined with the target text area among text areas having a color within a preset color range, and refers to the combined list to thereby generate the target text area. and the inclusive image including a text region in which the determining means has determined, the image compression apparatus according to any one of claims 1 to 6, characterized in that to generate the information related to the inclusive image.   The image compression apparatus according to claim 1, wherein the compression unit performs MMR compression on an included image and / or a text region having one color.   The image compression apparatus according to claim 1, wherein the compression unit performs lossless compression on a text region having a plurality of colors. Further, in the color document image, a background image generation means for generating a background image in which a text area is filled with a predetermined color;
The image compression apparatus according to any one of claims 1 to 9, characterized in that it comprises a base image compressing means for compressing the base image. The underlying image generating means, the position of the text area, above with reference to text information including a text area size of the image compression apparatus according to claim 1 0, characterized in that identifying the text area. The underlying image generating means, image compression apparatus according to claim 1 0, characterized in that to fill the text area by the average value of the pixels of the text area near the color document image. The underlying image compression means, the image compression apparatus according to claim 1 0, characterized in that performs JPEG compression on the underlying image. An image compression method performed by an image compression apparatus for compressing a color document image,
An extraction step of extracting a color of a text region included in the color document image;
A generating step for generating an inclusion image including a text region having a color within a preset color range in the text region;
A compression step of compressing the inclusion image and / or the text region ,
The generating step includes a determining step of determining a text region to be combined with a target text region among text regions having a color within a preset color range,
Generating an inclusion image including the text region of interest and the text region determined in the determination step, and information about the inclusion image;
In the determination step,
The first compressed size obtained by the sum of the compressed size estimated when the target text area or the inclusion image including the target text area is compressed and the compressed size estimated when the text area is compressed When,
Obtaining a second compressed size estimated when the attention image area or an inclusion image including the attention text area and an inclusion image including the text area are compressed;
An image compression method comprising combining the text area with the text area of interest when the second compressed size is smaller than the first compressed size . The computer program for causing to function as each unit included in the image compression apparatus according to any one of claims 1 to 1 3. A computer-readable storage medium storing the program according to claim 15 .

Images