JP2004128881A - Image compression apparatus, image expansion apparatus, and method for them, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image compression apparatus which generates an excellent decoded image without degrading image quality of a compression object and efficiently compresses an image at low cost performance, and also to provide an image expansion apparatus, a method for them, and a program. <P>SOLUTION: A binary section 102 binarizes a multi-value image 101. An area particularizing section A104 particularizes a position of a character area in a binary image 103. An area particularizing section B105 particularizes a position of a unit character area in the character area. A reduction section 114 reduces a processing section image 101. A representative color extract section 110 extracts a representative color of the unit character area on the basis of position information of the unit character area and a reduced multi-value image 118. A character area filling-in section 113 fills in the character area of the reduced multi-value image by its peripheral color on the basis of the position information of the character area. A JPEG compression section 108 compresses a partial binary image 107 corresponding to the character area. An MMR compression section 108 compresses the partial binary image 107 corresponding to the character area. Compression data 117 including the position information, compression codes generated by each compression section, and representative color information of the unit character area are outputted. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an image compression device for compressing an input multi-valued image, an image decompression device for decompressing compressed data generated by the image compression device, a method thereof, and a program.
[0002]
[Prior art]
2. Description of the Related Art In recent years, documents have been digitized due to the spread of scanners. If the digitized document is stored in a full-color bitmap format, for example, in the case of A4 size, the reading resolution is 300 dpi and the resolution becomes about 24 Mbytes, and the required memory becomes enormous. Such a large amount of data cannot be said to be a size suitable for being attached to a mail and transmitted.
[0003]
Therefore, compression of a full-color image is usually performed, and JPEG is known as a compression method. JPEG is very effective for compressing natural images such as photographs, but JPEG compression of character parts causes image degradation called mosquito noise. Therefore, conventionally, the input image is divided into a character region and a photograph region, and the character region portion is binarized and then subjected to MMR compression, and the photograph region portion is subjected to JPEG compression to improve the quality of the character region. There was a way to express full-color documents with a small data size while keeping them.
[0004]
According to this method, at the time of decompression, a white portion of a binary image is transmitted through a JPEG image, and a black portion is represented by characters with a representative color. There is also a method for further improving the compression ratio by filling a character area with a color near the character before JPEG compression. In addition, in order to realize efficient compression and good image reproduction at the time of decompression, a technique for extracting a representative color of a character area is essential.
[0005]
[Problems to be solved by the invention]
However, in the related art, in order to extract a representative color of a character area, an input multi-value image and a binary image obtained by binarizing the input multi-value image are required. It was necessary to temporarily hold the data in a memory such as the above. For example, when the input multi-valued image is A4 size, a read resolution of 300 dpi and a temporary memory of about 24 Mbytes are required. In particular, when a conventional compression technique is incorporated into a multifunction peripheral or a scanner having an image reading function, a reduction in the amount of memory is desired in terms of cost.
[0006]
The present invention has been made in order to solve the above-described problems, and it is possible to generate a good restored image and perform image compression efficiently and at a low cost performance without deteriorating the image quality of a compression target. It is an object of the present invention to provide an image compression device, an image decompression device, a method thereof, and a program that can be used.
[0007]
[Means for Solving the Problems]
An image compression apparatus according to the present invention for achieving the above object has the following configuration. That is,
An image compression device for compressing an input multi-valued image,
Binarizing means for binarizing the multi-valued image;
First area specifying means for specifying a position of a character area in the binary image binarized by the binarizing means;
A second area specifying means for specifying a position of a unit character area in the character area specified by the first area specifying means;
Reducing means for reducing the multi-valued image,
A representative color extracting unit that extracts a representative color of the unit character region based on the position information of the unit character region specified by the second region specifying unit and the reduced multivalued image;
Filling means for filling a unit character area in the character area of the reduced multi-value image with its surrounding color based on the position information of the character area specified by the first area specifying means;
First compression means for compressing the fill-in reduced multivalued image generated by the fill-in means;
A second compression unit that compresses a partial binary image corresponding to the character region specified by the first region specification unit;
Position information specified by the first and second area specifying means, first and second compressed codes generated by the first and second compression means, and a unit character area extracted by the representative color extraction means And output means for outputting compressed data including the representative color information.
[0008]
Also preferably, the representative color extracting unit refers to the color information of the reduced multi-valued image using the position information of the unit character region specified by the second region specifying unit, thereby obtaining the unit character region. Average color calculating means for calculating the average color of
First histogram creating means for creating a first histogram from the average color of the unit character area generated by the average color calculating means;
Extracting means for extracting a candidate representative color that is a candidate for a representative color of the unit character area based on the first histogram;
A color assigning unit that assigns the candidate representative color to the representative color of the unit character based on the candidate representative color extracted by the extracting unit and the average color of the unit character area.
[0009]
Preferably, the binarizing means includes a second histogram creating means for creating a second histogram based on the multi-valued image,
And a binarization threshold value calculating means for calculating a binarization threshold value for binarizing the multi-valued image based on the second histogram.
[0010]
Preferably, the first area specifying means specifies an area group specified by performing contour line tracing of pixels having a predetermined value in the binary image as a character area or a natural image area, respectively. Position information, size information, and attribute information indicating the type of each area of the group are specified.
[0011]
Preferably, the second area specifying means specifies, as a unit character area, a set of pixels having a predetermined value in the character area specified by the first area specifying means. Specify size information.
[0012]
Also, preferably, the average color calculating means uses the position information of the unit character area specified by the second area specifying means, and uses the reduced multi-value corresponding to a pixel having a predetermined value in the unit character area. The average color of the unit character area is calculated by referring to the color information of the image.
[0013]
Preferably, the first histogram creating unit uses the average color of the unit character area calculated by the average color calculating unit to set a color space histogram in the character area including the unit character area as a first histogram. create.
[0014]
Preferably, the extraction unit extracts candidate representative colors that are candidates for the representative color of the unit character area in order from a color distribution having a high output frequency in the color space histogram.
[0015]
Preferably, the color allocating unit calculates a distance in a color space between the candidate representative color extracted by the extracting unit and an average color of the unit character area, and the distance is set within a predetermined range. In some cases, the candidate representative color is assigned as a representative color of the unit character area.
[0016]
Preferably, the reduction unit reduces the multi-valued image by performing resolution conversion of the multi-valued image using a resolution conversion parameter.
[0017]
Also, preferably, the filling means includes the reduced multi-value corresponding to a position corresponding to a reduction degree set by the reducing means, based on position information of the character area specified by the first area specifying means. The unit character area in the character area of the image is filled with surrounding colors.
[0018]
Preferably, the first compression unit is based on JPEG compression.
[0019]
Preferably, the second compression unit is based on MMR compression.
[0020]
Preferably, the apparatus further comprises third compression means for reversibly compressing the compressed data.
[0021]
An image decompression device according to the present invention for achieving the above object has the following configuration. That is,
An image decompression device for decompressing compressed data,
A first compression code for compressing a character area in a binary image obtained by binarizing a multi-valued image, and a unit character area in the character area of the reduced multi-valued image of the multi-valued image filled with surrounding colors. Input means for inputting compressed data including a second compressed code obtained by compressing the reduced multi-valued image, position information of the character area, and representative color information of a unit character area in the character area;
First decompression means for decompressing the first compression code in the compressed data;
Second decompression means for decompressing the second compression code in the compressed data;
Enlarging means for enlarging the reduced multivalued image expanded by the second expanding means;
A synthesizing unit that synthesizes the binary image expanded by the first expansion stage and the multi-value image expanded by the expansion unit based on the position information and the representative color in the compressed data.
[0022]
An image compression method according to the present invention for achieving the above object has the following configuration. That is,
An image compression method for compressing an input multi-valued image,
A binarization step of binarizing the multi-valued image;
A first area specifying step of specifying a position of a character area in the binary image binarized in the binarizing step;
A second area specifying step of specifying the position of the unit character area in the character area specified in the first area specifying step;
A reducing step of reducing the multi-valued image,
A representative color extracting step of extracting a representative color of the unit character area based on the position information of the unit character area specified in the second area specifying step and the reduced multi-valued image;
A filling step of filling a unit character area in the character area of the reduced multi-valued image with its surrounding color based on the position information of the character area specified in the first area specifying step;
A first compression step of compressing the fill-in reduced multivalued image generated in the fill-in step;
A second compression step of compressing a partial binary image corresponding to the character area specified in the first area specification step;
The position information specified in the first and second area specifying steps, the first and second compressed codes generated in the first and second compression steps, and the unit character area extracted in the representative color extraction step Outputting compressed data including the representative color information.
[0023]
An image decompression method according to the present invention for achieving the above object has the following configuration. That is,
An image decompression method for decompressing compressed data,
A first compression code for compressing a character area in a binary image obtained by binarizing a multi-valued image, and a unit character area in the character area of the reduced multi-valued image of the multi-valued image filled with surrounding colors. An input step of inputting compressed data including a second compressed code obtained by compressing the reduced multi-valued image, position information of the character area, and representative color information of a unit character area in the character area;
A first decompression step of decompressing the first compression code in the compressed data;
A second decompression step of decompressing the second compression code in the compressed data;
An enlarging step of enlarging the reduced multivalued image expanded in the second expanding step;
A synthesizing step of synthesizing the binary image decompressed in the first decompression stage and the multi-value image expanded in the enlarging step based on the position information and the representative color in the compressed data.
[0024]
A program according to the present invention for achieving the above object has the following configuration. That is,
A program for realizing image compression for compressing an input multi-valued image,
Program code for a binarization step of binarizing the multi-valued image;
A program code for a first area specifying step for specifying a position of a character area in the binary image binarized in the binarizing step;
A program code for a second area specifying step for specifying the position of the unit character area in the character area specified in the first area specifying step;
Program code for a reduction step of reducing the multi-valued image,
A program code for a representative color extracting step of extracting a representative color of the unit character area based on the position information of the unit character area specified in the second area specifying step and the reduced multi-valued image;
A program code for a filling process of filling a unit character region in the character region of the reduced multi-valued image with its surrounding color based on the position information of the character region specified in the first region specifying step;
A program code of a first compression step of compressing the fill-in reduced multi-value image generated in the padding step;
Program code for a second compression step of compressing a partial binary image corresponding to the character area specified in the first area specification step;
The position information specified in the first and second area specifying steps, the first and second compressed codes generated in the first and second compression steps, and the unit character area extracted in the representative color extraction step And a program code of an output step of outputting compressed data including the representative color information.
[0025]
A program according to the present invention for achieving the above object has the following configuration. That is,
A program for realizing image expansion for expanding compressed data,
A first compression code for compressing a character area in a binary image obtained by binarizing a multi-valued image, and a unit character area in the character area of the reduced multi-valued image of the multi-valued image filled with surrounding colors. A second compression code for compressing the reduced multi-valued image, position information of the character area, and program code of an input step of inputting compressed data including representative color information of a unit character area in the character area;
Program code for a first decompression step for decompressing the first compression code in the compressed data;
Program code for a second decompression step for decompressing the second compression code in the compressed data;
Program code for an enlarging step for enlarging the reduced multivalued image expanded in the second expanding step;
And a program code for a synthesizing step of synthesizing the binary image expanded in the first expansion stage and the multi-valued image expanded in the expansion step based on the position information and the representative color in the compressed data. .
[0026]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
[0027]
It should be noted that the relative arrangement of each component of the image compression apparatus and the image decompression apparatus of the embodiment described below, and the mathematical formulas and numerical values used for each processing are not limited to the scope of the present invention unless otherwise specified. Is not intended to be limited only to them.
[0028]
FIG. 1 is a diagram showing a schematic configuration of an image compression device according to an embodiment of the present invention.
[0029]
In FIG. 1, the solid line indicates the flow and input of an image, and the dotted line indicates the flow and input of information.
[0030]
An input image 101, which is a color multi-valued image, is binarized by a binarization unit 102, and a binary image 103 is created. The area specifying unit A104 receives the binary image 103, detects a character area by, for example, tracing the contour of a pixel (for example, a black pixel) having a predetermined value, and creates the character area coordinates 106. The character area coordinates 106 are information indicating the position (coordinates) and size of the character area. It is needless to say that, by specifying the character area by the area specifying unit A104, the position and size of a natural image area indicating a natural (gradation) image such as a photograph or an illustration are also specified. Further, attribute information (characters and images) for specifying the type of each area is separately generated.
[0031]
The area specifying unit B105 receives the binary image 103 and the character area coordinates 106 created by the area specifying unit A104, and specifies the position and size of each character (unit character area) in the character area. For simplicity, in the first embodiment, information on the position and size of each unit character area is also added to the character area coordinates 106. Further, a binary image (partial binary image 107) for each character area is created based on the character area coordinates 106 created by the area specifying unit A104. Further, the multi-valued image 112 is reduced by the reduction unit 114, and a reduced multi-valued image 118 is created.
[0032]
The reduction unit 114 performs resolution conversion of the fill-in-the-blank image based on the set resolution conversion parameters. The reduction unit 114 may appropriately control the resolution conversion parameter based on the spatial frequency of the multivalued image to be processed. For example, when the high-frequency component of the multi-valued image is small, the resolution conversion parameter may be controlled so as to convert to a lower resolution than when the high-frequency component is large.
[0033]
The representative color extraction unit 110 receives the partial binary image 107, the character area coordinates 106, and the reduced multi-valued image 118 as input, and associates the black area of the partial binary image 107 with the reduced multi-valued image 118 to position the character area. Each character representative color 111 of each unit character area in the middle is calculated. The details of this process will be described later. Note that the multi-valued image 112 is the same as the input image 101.
[0034]
The character area filling unit 113 receives the reduced multi-valued image 114 and the character area coordinates 106 as input, and fills each character area or unit character area on the reduced multi-valued image 118 with its surrounding color (replace with the surrounding color). Process).
[0035]
After the above processing, each of the partial binary images 107 is sequentially compressed by the MMR compression unit 108 as a compression code A109. Further, the multi-valued image filled in by the character area filling unit 113 is JPEG-compressed by the JPEG compression unit 115 as a compression code B116. Note that, instead of the MMR compression unit 108, binary image compression other than MMR compression, for example, JBIG compression, MR compression, MH compression, or the like may be used.
[0036]
In this way, the compressed data 117 is created by combining the data group of the character area coordinates 106, the compression code A109, each character representative color 111, and the compression code B116 obtained from each component. The compressed data 117 may be further losslessly compressed to PDF or the like.
[0037]
If there is no character area in the input image 101, the compressed data 117 is only the compressed code B116.
[0038]
The hardware configuration for realizing the image compression device and an image decompression device that decompresses the compressed data generated by the image compression device described later is realized by a general-purpose computer such as a personal computer, for example. Further, this general-purpose computer has, as standard components, a CPU, a RAM, a ROM, a hard disk, an external storage device, a network interface, a display, a keyboard, a mouse, and the like.
[0039]
The image compression device and the image decompression device may be realized as dedicated hardware realized as an expansion card for such a general-purpose computer.
[0040]
Further, specific examples of an apparatus equipped with the image compression apparatus and the image decompression apparatus include a multifunction peripheral having a network communication function (an apparatus having a copy, printer, scanner, facsimile function, etc.), a color scanner, a color facsimile, and the like. No.
[0041]
Next, a binarization process performed by the binarization unit 102 will be described with reference to FIGS.
[0042]
FIG. 2 is a diagram illustrating an example of an input image according to the embodiment of the present invention. FIG. 3 is a flowchart illustrating a binarization process executed by the binarization unit according to the embodiment of the present invention.
[0043]
In FIG. 2, it is assumed that an input image 201 is a color multivalued image, an area 203 is red, an area 204 is blue, other characters are black, and an area 205 is an arbitrary plurality of colors.
[0044]
As an example, when the input image 101 is to be RGB data (24 bits) after being read by a scanner, first, in step S301, brightness conversion is performed by the following conversion formula to create a brightness image.
[0045]
Y = 0.299 × R + 0.587 × G + 0.114 × B
Next, in step S302, an entire histogram of the luminance image is created. Here, FIG. 6 shows an example of the histogram. In FIG. 6, the horizontal axis represents the luminance level of the Y signal from 0 to 255, and the vertical axis represents its appearance frequency. In the case of FIG. 6, 601 indicates the distribution of characters and images, and 602 indicates the distribution of the background.
[0046]
Next, in step S303, an optimal binarization threshold T is calculated. However, the method of calculating the binarization threshold T here is not particularly limited. In FIG. 6, for example, an intermediate point 603 between the luminance levels at the vertices of the distribution 601 and the distribution 602 is set as the binarization threshold T.
[0047]
Finally, in step S304, the luminance image is binarized based on the binarization threshold T.
[0048]
Through the above processing, the binary image 103 of FIG. 1 is created. A binary image obtained by binarizing the multi-valued image 201 in FIG. 2 is, for example, a binary image 701 in FIG.
[0049]
Next, a process executed by the area specifying unit A104 will be described with reference to FIG.
[0050]
FIG. 4 is a flowchart showing a process executed by the area specifying unit A104 according to the embodiment of the present invention.
[0051]
First, in step S401, contour tracing is performed with reference to black pixels of the binary image 103. Next, in step S402, the inside of the traced outline is further traced, and based on the result of the trace, a character region and its position and size are specified from the region within the outline. In addition, an area other than the character area is specified as a natural image area.
[0052]
Through the above processing, the attributes indicating the position and size of the character area and the natural image area and the type of the area are specified. Here, FIG. 8 shows a result obtained by performing the processing of the area specifying unit A104 on the binary image 701 of FIG. In FIG. 8, four regions are specified in total, 801 to 803 are specified as character regions, and 804 is specified as a natural image region.
[0053]
Next, a process executed by the area specifying unit B105 will be described with reference to FIG.
[0054]
FIG. 5 is a flowchart showing a process executed by the area specifying unit B105 according to the embodiment of the present invention.
[0055]
The processing described in FIG. 5 is sequentially performed on the character area group specified by the area specifying unit A104, in the example of FIG. 8, the character areas 801 to 803.
[0056]
First, in step S501, it is determined whether there is an unprocessed character area. If there is no unprocessed character area (NO in step S501), the process ends. On the other hand, if there is an unprocessed character area (YES in step S501), the process advances to step S502.
[0057]
Next, in step S502, a set of pixels having a predetermined value (black pixel) of the binary image is regarded as a unit character, and the position of the unit character area is specified. Here, FIG. 9 shows a result obtained by performing the processing of the area specifying unit B105 on the character area 801 of FIG. In FIG. 9, a total of five unit character areas 902 to 906 are specified. Thereafter, the processing is similarly performed for the character areas 802 and 803 in FIG.
[0058]
As described above, the area information (position and size) of the character area / unit character area specified by the area specifying unit A104 and the area specifying unit B105 is used as the character area coordinates 106, for example, in the memory ( (For example, RAM).
[0059]
Next, the processing executed by the representative color extracting unit 110 will be described with reference to the flowcharts of FIGS. 10 and 11 and the explanatory diagrams of FIGS.
[0060]
FIG. 10 is a flowchart illustrating a process executed by the representative color extracting unit according to the embodiment of the present invention.
[0061]
As described above, the representative color extraction unit 110 uses the character area coordinates 106, the partial binary image 107, and the reduced multi-value image 118 as inputs. It is assumed that the partial binary image 107 has been obtained by cutting out the character area of the binary image 103 and storing it in another memory such as a work memory. The reduced multi-valued image 118 is obtained by reducing the multi-valued image 112 by the reduction unit 114, and is stored in another memory such as a work memory.
[0062]
In the present embodiment, the description will be made assuming that the reduction degree of the reduction unit 114 is 50%. In this case, the work memory for the reduced multi-valued image can be reduced in memory capacity to 1/4 that of the work memory for the multi-valued image before the reduction. However, the degree of reduction is not limited to 50%, and it goes without saying that an arbitrary reduction rate can be set according to the application and purpose.
[0063]
Here, it is assumed that the character area 801 in FIG. 9 is used as a specific example of the processing target.
[0064]
Since the representative color extracting unit 110 sequentially processes all the character areas, that is, all the partial binary images 107, first, in step S <b> 1001, the unprocessed The presence or absence of a character area is determined. If there is no unprocessed character area (NO in step S1001), the process ends. If there is an unprocessed character area (YES in step S1001), the process advances to step S1008.
[0065]
Next, in step S1008, 1 is set to the representative color number n of the character area. This means "extract the first color". Next, in step S1002, while referring to the character area coordinates 106 and referring to the color (RGB) of the reduced multi-valued image 114 corresponding to the position of the black pixel of the partial binary image 107, each unit character The average color of RGB is calculated. This calculation example will be specifically described with reference to FIGS. 12A and 12B.
[0066]
In FIG. 12A, reference numeral 201 denotes a multivalued image, reference numeral 701 denotes a binary image after binarization, and reference numeral 203 denotes a reduced multivalued image after reduction. The image end coordinates of the binary image 701 are p1 (0, 0), p2 (w, 0), p3 (0, h), and p4 (w, h). On the other hand, since the reduced image 203 is reduced by 50% vertically and horizontally, the image end coordinates are upper left p9 (0, 0), upper right p10 (w / 2, 0), lower left p11 (0, h / 2), It can be expressed as lower right p12 (w / 2, h / 2).
[0067]
Here, w is the width of the multi-valued image 201, and h is the height of the multi-valued image 201.
[0068]
Reference numeral 1201 denotes an enlarged view of the unit character area “C” on the binary image 701, and reference numeral 1202 denotes an enlarged view of the unit character area “C” on the reduced image 203. The unit character coordinates in the enlarged view 1201 are upper left p5 (x, y), upper right p6 (x + tw, y), lower left p7 (x, y + th), and lower right p8 (x + tw, y + th). On the other hand, the enlarged view 1202 shows the upper left p13 (x / 2, y / 2), the upper right p14 ((x + tw) / 2, y / 2), the lower left p15 (x / 2, (y + th) / 2), and the lower right. p16 ((x + tw) / 2, (y + th) / 2).
[0069]
However, the information of x, y, tw, and th is obtained from the character area coordinates 106. In calculating the RGB average color for each unit character in step S1002, processing at the pixel level is performed. A specific example of this processing will be described with reference to FIG. 12B.
[0070]
In FIG. 12B, reference numerals 1701 and 1702 denote the pixel distributions of the enlarged views 1201 and 1202, respectively.
[0071]
Here, for example, it can be seen that the pixel of the pixel distribution 1702 that corresponds in position to the black pixel of p17 (x + 2, y + 2) of the pixel distribution 1701 is p19 ((x + 2) / 2, (y + 2) / 2). . Then, the RGB values of p19 ((x + 2) / 2, (y + 2) / 2) are referred to. Further, since the pixel distribution 1702 is reduced at a reduction degree of 50%, the same RGB value of p19 ((x + 2) / 2, (y + 2) / 2) is referred to at p18 (x + 3, y + 3).
[0072]
Hereinafter, the pixel values (RGB values) of the pixel distribution 1702 corresponding to all the black pixels in the pixel distribution 1701 are referred to, and the average value of the RGB values of the unit character is calculated using the pixel values obtained by the reference. I do.
[0073]
That is, by using the position information of the unit character area on the binary image 701, referring to the color information (RGB values) corresponding to the position on the reduced multi-valued image 203 according to the degree of reduction of the reducing unit 114, The average color of the unit character area on the binary image 701 is calculated.
[0074]
Next, in step S1003, the RGB average color of each unit character area calculated above is converted into a color space such as YCbCr that can be expressed by luminance and color difference. The color space may be another color space such as Lab or Yuv, but here, the description will be made with YCbCr. Further, the conversion formula from RGB to YCbCr is omitted.
[0075]
For example, when the above five unit character areas 902 to 906 are subjected to YCbCr conversion,
Unit character area 902: (Y1, Cb1, Cr1)
Unit character area 903: (Y2, Cb2, Cr2)
Unit character area 904: (Y3, Cb3, Cr3)
Unit character area 905: (Y4, Cb4, Cr4)
Unit character area 906: (Y5, Cb5, Cr5)
It becomes.
[0076]
Next, in step S1004, histogram initialization (histogram clear) is performed in preparation for the histogram creation in step S1005. Next, in step S1005, a two-dimensional histogram of the color differences Cb and Cr is created. This is shown in FIG. 13 are plot 1301 (Cb1, Cr1) → plot plot 1302 (Cb2, Cr2) of unit character area 902 → plot plot 1303 (Cb3, Cr3) of unit character area 903 → unit character A plot plot 1304 (Cb4, Cr4) of the area 904 → a plot plot 1305 (Cb5, Cr5) of the unit character area 905 → a plot of the unit character area 906. The plots 1301, 1302 and 1305 are spaced apart from the plots 1303 and 1304 in order to represent the difference in the colors shown in the multi-valued image 201 in FIG. Or the difference between the plots 1303 and 1304 indicates the variation in the input image read by the scanner or the variation due to the subsequent processing (for example, JPEG compression).
[0077]
In the present embodiment, a two-dimensional histogram is created for speeding up processing, but a three-dimensional histogram of Y, Cb, and Cr may be created for improving accuracy.
[0078]
Next, in step S1006, representative color extraction and color assignment of each unit character area are performed. The details of this process will be described later.
[0079]
Next, in step S1007, it is determined whether there is a unit character whose color has not been allocated in step S1006. If there is an unassigned unit character (YES in step S1007), the flow advances to step S1009 to add 1 to the representative color number n to initialize the previously created histogram (step S1004). A histogram is created only for characters (step S1005). Hereinafter, similarly, in step S1007, the process is continued until there is no unit character to which a color has not been assigned. If there is no unassigned unit character (NO in step S1007), the process returns to step S1001, and if there is another unprocessed character area, the same processing is performed.
[0080]
Next, details of the process of extracting the representative color of each unit character area and assigning colors in step S1006 will be described with reference to FIG.
[0081]
FIG. 11 is a flowchart showing details of the representative color extraction and color assignment processing in step S1006 according to the embodiment of the present invention.
[0082]
First, in step S1101, a candidate representative color C ′ (n) that is a candidate for a representative color of a unit character area is extracted. Note that n is the number of representative colors, which is the same as n used in FIG. Further, the candidate representative color C ′ (n) is a structure having information of Cb and Cr. n) _r.
[0083]
A specific method of extracting candidate representative colors will be described with reference to FIG.
[0084]
A square surrounded by a dotted line shown in FIG. 14 indicates a preset area, and it is assumed that one representative color (Cb, Cr) is set in each area. Although the size of the area is arbitrary, it is desirable to set the area in consideration of a trade-off between accuracy and processing efficiency. In a case where each of YCbCr is 8 bits, in the present embodiment, the vertical and horizontal widths of the area are set to 3 bits. However, in FIG. 14, the area is illustrated in a large size for the sake of simplicity. Here, each area is scanned in the Cb-Cr direction, and the maximum value of the number of plots in the area is obtained. Here, since the maximum number of plots 3 is obtained in the area 1, 1401 (black triangle）) is obtained as the candidate representative color.
[0085]
Next, in step S1102, 1 is set for the number m of unit characters. Next, in step S1103, it is determined whether or not m is equal to or less than M. If it is less than or equal to M (YES in step S1103), the process advances to step S1104. On the other hand, if it is larger than M (NO in step S1103), the process ends.
[0086]
Here, M is the number of unit characters (total number) in the character area to be processed. For example, in the character area 801 in FIG. 9, M = 5. In this case, when m = 1, the condition of step S1103 is YES, and the process proceeds to step S1104. Then, when m = 6, the condition of step S1103 is NO, and the process exits the loop of step S1006.
[0087]
In step S1104, it is determined whether the unit character color C (m), which is the average color of the unit character area, has not been determined. If it is not determined (NO in step S1104), the process advances to step S1108. On the other hand, if it has not been determined (YES in step S1104), the flow advances to step S1105.
[0088]
Here, the unit character color C (m) is a structure having information of Cb and Cr. ) _R. Since no value is initially assigned to the unit character color C (m), C (m) is undetermined and the process proceeds to step S1105.
[0089]
Next, in step S1105, a distance (Distance) in the color space between the candidate representative color C ′ (n) and the unit character color C (m) is calculated.
[0090]
Next, in step S1106, it is determined whether or not Distance ≦ value with respect to a preset value value. If Distance ≦ value is not satisfied (NO in step S1106), the flow advances to step S1108. On the other hand, if Distance ≦ value is satisfied (YES in step S1106), the flow advances to step S1107.
[0091]
In step S1107, the candidate representative color C ′ (n) is set as the unit character color C (m) (the representative color of the unit character area). That is, C '(n) _b is substituted for C (m) _b, and C' (n) _r is substituted for C (m) _r. Here, the value of the value can be arbitrarily adjusted.
[0092]
Then, in step S1108, 1 is added to the number of unit characters m, and the process returns to step S1103.
[0093]
A specific example of the processing in steps S1105 and S1106 in FIG. 11 will be described with reference to FIG. Here, since it is assumed that n = 1 and m = 1, the candidate representative color C ′ (1) is a plot 1401 and the unit character color C (1) is a plot 1301. In the above-described steps S1105 and S1106, whether the plot 1301 of the unit character color C (1) exists in the dotted circle 1500 centered on the plot 1401 that is the candidate representative color C ′ (1) and having a radius of value Has been determined.
[0094]
Here, since the plot 1301 exists in the dotted circle 1500, the candidate representative color C ′ (1) (black or an approximate color of black in this embodiment) is changed to the representative color C corresponding to the unit character area 902. Set as (1). Similarly, when m = 2 and 5, the plots 1302 and 1305 satisfy the condition of step S1106, so that the candidate representative colors C ′ (1) are replaced with the representative colors C (2) and C (2) of the unit character areas 903 and 906, respectively. Set as C (5).
[0095]
On the other hand, when m = 3, 4, since the condition of step S1106 is not satisfied, the process leaves the loop of step S1006 and performs the next representative color extraction while leaving C (3) and C (4) undecided. Then, in the subsequent processing, the candidate representative colors C ′ (2) (blue or an approximate color of blue in this embodiment) are changed to the representative colors C (3) and C (3) corresponding to the unit character areas 904 and 905, respectively. 4).
[0096]
As described above, in the flowcharts shown in FIGS. 10 and 11, the representative color extraction and the color assignment of each unit character area are performed. The extracted representative colors are stored in the compressed data 117 as the respective character representative colors 111 in FIG.
[0097]
Next, the processing executed by the character area filling unit 113 will be described with reference to FIG.
[0098]
FIG. 16 is a flowchart showing processing executed by the character area filling unit according to the embodiment of the present invention.
[0099]
The character area filling unit 113, the reduced multi-valued image 118, the partial binary image 107, and the character area coordinates 106 are input, and pixels corresponding to characters in the reduced multi-valued image 118 are filled with the surrounding background color. As a result, the compression ratio of the JPEG compression unit 115 is improved.
[0100]
First, in step S1801, to perform processing for each character area, it is determined whether there is an unprocessed character area. If there is no unprocessed character area (NO in step S1801), the processing ends. On the other hand, if there is an unprocessed character area (YES in step S1801), the flow advances to step S1802 to refer to the color of the reduced multi-valued image 118 corresponding to the position of the white pixel of the partial binary image 107 to obtain the character. The average value of the background color in the area is calculated. The method of coordinating the coordinates of the partial binary image 107 and the reduced multi-valued image 118 is the same as the method described for the representative color extracting unit 110, and thus the details are omitted. Next, in step S1803, the calculated average value of the background color is assigned to the character area of the reduced multi-valued image 118, that is, the character area of the reduced multi-valued image 118 or the unit character in the character area using the calculated background color. Fill the area.
[0101]
Next, an image decompression device that decompresses the compressed data 117 will be described with reference to FIG.
[0102]
FIG. 17 is a diagram showing a schematic configuration of the image decompression device according to the embodiment of the present invention.
[0103]
The MMR decompression unit 1603 receives the compression code A109, performs an MMR decompression process, and creates a binary image 1604. The JPEG decompression unit 1609 receives the compression code B116, performs JPEG decompression processing, and further performs expansion processing in the expansion unit 1610, thereby creating a multi-valued image 1611. The combining unit 1605 assigns each character representative color 111 to each black pixel of the corresponding unit character area in the binary image 1604 while referring to the character area coordinates 106, and places the binary image on the multi-valued image 1611. indicate. At this time, the white pixels of the binary image 1604 are transmitted through the multi-value image 1611.
[0104]
In this way, the image decompression device of FIG. 17 can decompress the compressed data 117 created by the image compression device of FIG. 1 and generate a decompressed image 1606 that is the final restored image.
[0105]
As described above, according to the present embodiment, after reducing the input multi-valued image and using the reduced multi-valued image to extract the representative color of the unit character area of the character area, at least the conventional technique is used. A work memory for temporarily storing an image to be processed can be significantly reduced while maintaining the representative color extraction accuracy. Further, since the number of processing steps can be realized by the same number of processing steps as in the past, cost performance can be improved.
[0106]
Further, a representative color in a binary image of a character area and a unit character area constituting the character area is extracted and assigned, and the representative color of each unit character area is generated as a part of the compressed data. Then, when a decompressed image is obtained by expanding the compressed data, the color of each unit character area in each character area is reproduced using this representative color. This makes it possible to reproduce a restored image in which the image quality of the input image is well maintained.
[0107]
Further, with respect to the character area, a compressed image with higher compression can be generated by using MMR compression instead of the conventionally used JPEG compression.
[0108]
Although the embodiment has been described in detail, the present invention may be applied to a system including a plurality of devices, or may be applied to an apparatus including a single device.
[0109]
According to the present invention, a software program (in the embodiment, a program corresponding to the flowchart shown in the drawings) for realizing the functions of the above-described embodiments is directly or remotely supplied to a system or an apparatus, and the computer of the system or the apparatus is supplied to the computer. Is also achieved by reading and executing the supplied program code.
[0110]
Therefore, since the functions of the present invention are implemented by computer, the program code installed in the computer also implements the present invention. That is, the present invention includes the computer program itself for realizing the functional processing of the present invention.
[0111]
In this case, any form of the program, such as an object code, a program executed by an interpreter, and script data supplied to the OS, is applicable as long as the program has the function of the program.
[0112]
As a recording medium for supplying the program, for example, a floppy (registered trademark) disk, hard disk, optical disk, magneto-optical disk, MO, CD-ROM, CD-R, CD-RW, magnetic tape, nonvolatile memory card , ROM, DVD (DVD-ROM, DVD-R) and the like.
[0113]
Other methods of supplying the program include connecting to a homepage on the Internet using a browser of a client computer, and downloading the computer program itself of the present invention or a file containing a compressed automatic installation function from the homepage to a recording medium such as a hard disk. Can also be supplied. Further, the present invention can also be realized by dividing the program code constituting the program of the present invention into a plurality of files and downloading each file from a different homepage. That is, the present invention also includes a WWW server that allows a plurality of users to download a program file for implementing the functional processing of the present invention on a computer.
[0114]
In addition, the program of the present invention is encrypted, stored in a storage medium such as a CD-ROM, distributed to users, and downloaded to a user who satisfies predetermined conditions from a homepage via the Internet to download key information for decryption. It is also possible to execute the encrypted program by using the key information and install the program on a computer to realize the program.
[0115]
The functions of the above-described embodiments are implemented when the computer executes the read program, and an OS or the like running on the computer executes a part of the actual processing based on the instructions of the program. Alternatively, all the operations are performed, and the functions of the above-described embodiments can be realized by the processing.
[0116]
Further, after the program read from the recording medium is written into a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer, the function expansion board or the A CPU or the like provided in the function expansion unit performs part or all of the actual processing, and the processing also realizes the functions of the above-described embodiments.
[0117]
【The invention's effect】
As described above, according to the present invention, it is possible to generate a good restored image, and an image compression apparatus capable of performing image compression without deteriorating the image quality of a compression target efficiently and at low cost performance, An image decompression device, a method thereof, and a program can be provided.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating a schematic configuration of an image compression device according to an embodiment of the present invention.
FIG. 2 is a diagram illustrating an example of an input image according to the embodiment of the present invention.
FIG. 3 is a flowchart illustrating a binarization process executed by a binarization unit according to the embodiment of the present invention.
FIG. 4 is a flowchart illustrating processing executed by an area specifying unit A according to the embodiment of the present invention.
FIG. 5 is a flowchart illustrating a process executed by an area specifying unit B according to the embodiment of the present invention.
FIG. 6 is a diagram illustrating an example of a histogram of an input image according to the embodiment of the present invention.
FIG. 7 is a diagram illustrating an example of a binary image according to the embodiment of the present invention.
FIG. 8 is a diagram illustrating an example of a processing result by an area specifying unit A according to the embodiment of the present invention.
FIG. 9 is a diagram illustrating an example of a processing result by a region specifying unit B according to the embodiment of the present invention.
FIG. 10 is a flowchart illustrating a process executed by a representative color extracting unit according to the embodiment of the present invention.
FIG. 11 is a flowchart illustrating details of representative color extraction and color allocation processing in step S1006 according to the embodiment of this invention.
FIG. 12A is a diagram for describing processing of a representative color extraction unit according to the embodiment of this invention.
FIG. 12B is a diagram for describing processing of the representative color extraction unit according to the embodiment of this invention.
FIG. 13 is a diagram for describing processing of a representative color extraction unit according to the embodiment of the present invention.
FIG. 14 is a diagram for explaining processing of a representative color extracting unit according to the embodiment of the present invention.
FIG. 15 is a diagram for describing processing of a representative color extraction unit according to the embodiment of this invention.
FIG. 16 is a flowchart illustrating a process executed by a character area filling unit according to the embodiment of the present invention.
FIG. 17 is a diagram illustrating a schematic configuration of an image decompression device according to an embodiment of the present invention.
[Explanation of symbols]
101 Input image 102 Binarization unit 103 Binary image 104 Area specifying unit A
105 Area specifying unit B
106 character area coordinates 107 partial binary image 108 MMR compression unit 109 compression code A
110 Representative color extraction unit 111 Each character representative color 112 Multi-valued image 113 Character area filling unit 114 Reduction unit 115 JPEG compression unit 116 Compression code B
117 Compressed data 118 Reduced multi-valued image 1603 MMR decompressor 1604 Binary image 1605 Combiner 1606 Decompressed image 1609 JPEG decompressor 1610 Enlarger 1611 Multi-valued image

Claims (19)

An image compression device for compressing an input multi-valued image,
Binarizing means for binarizing the multi-valued image;
First area specifying means for specifying a position of a character area in the binary image binarized by the binarizing means;
A second area specifying means for specifying a position of a unit character area in the character area specified by the first area specifying means;
Reducing means for reducing the multi-valued image,
A representative color extracting unit that extracts a representative color of the unit character region based on the position information of the unit character region specified by the second region specifying unit and the reduced multivalued image;
Filling means for filling a unit character area in the character area of the reduced multi-value image with its surrounding color based on the position information of the character area specified by the first area specifying means;
First compression means for compressing the fill-in reduced multivalued image generated by the fill-in means;
A second compression unit that compresses a partial binary image corresponding to the character region specified by the first region specification unit;
Position information specified by the first and second area specifying means, first and second compressed codes generated by the first and second compression means, and a unit character area extracted by the representative color extraction means Output means for outputting compressed data including the representative color information. The representative color extracting means calculates the average color of the unit character area by referring to the color information of the reduced multi-valued image using the position information of the unit character area specified by the second area specifying means. Means for calculating average color,
First histogram creating means for creating a first histogram from the average color of the unit character area generated by the average color calculating means;
Extracting means for extracting a candidate representative color that is a candidate for a representative color of the unit character area based on the first histogram;
2. The image processing apparatus according to claim 1, further comprising: a color assigning unit that assigns the candidate representative color to the representative color of the unit character based on the candidate representative color extracted by the extracting unit and an average color of the unit character area. An image compression device according to claim 1. A second histogram creating unit that creates a second histogram based on the multi-valued image;
3. The image compression apparatus according to claim 2, further comprising: a binarization threshold calculating unit configured to calculate a binarization threshold for binarizing the multi-valued image based on the second histogram. The first area specifying means specifies an area group specified by performing contour line tracing of a pixel having a predetermined value in the binary image, as a character area or a natural image area, respectively. 2. The image compression device according to claim 1, wherein attribute information indicating position information, size information, and a type thereof is specified. The second region specifying unit specifies a set of pixels having a predetermined value in the character region specified by the first region specifying unit as a unit character region, and specifies position information and size information of the unit character region. The image compression apparatus according to claim 1, wherein: The average color calculating means uses the position information of the unit character area specified by the second area specifying means to calculate color information of the reduced multivalued image corresponding to a pixel having a predetermined value in the unit character area. 3. The image compression apparatus according to claim 2, wherein an average color of the unit character area is calculated by referring to the reference color. The first histogram creating means creates a color space histogram in a character area including the unit character area as a first histogram using the average color of the unit character area calculated by the average color calculating means. The image compression apparatus according to claim 2, wherein 8. The image compression method according to claim 7, wherein the extraction unit extracts candidate representative colors that are candidates for the representative color of the unit character area in order from a color distribution having a high output frequency in the color space histogram. apparatus. The color allocating unit calculates a distance in a color space between the candidate representative color extracted by the extracting unit and the average color of the unit character area, and when the distance is within a predetermined range, 9. The apparatus according to claim 8, wherein a candidate representative color is assigned as a representative color of the unit character area. 2. The image compression apparatus according to claim 1, wherein the reduction unit reduces the multi-valued image by performing resolution conversion on the multi-valued image using a resolution conversion parameter. The fill-in-the-blank means includes a character area of the reduced multi-valued image corresponding to a position corresponding to the reduction degree set by the reducing means, based on the position information of the character area specified by the first area specifying means. 2. The image compression apparatus according to claim 1, wherein the unit character area is filled with surrounding colors. The image compression apparatus according to claim 1, wherein the first compression unit is based on JPEG compression. The image compression apparatus according to claim 1, wherein the second compression unit is based on MMR compression. 3. The image compression apparatus according to claim 1, further comprising: third compression means for reversibly compressing the compressed data. An image decompression device for decompressing compressed data,
A first compression code for compressing a character area in a binary image obtained by binarizing a multi-valued image, and a unit character area in the character area of the reduced multi-valued image of the multi-valued image filled with surrounding colors. Input means for inputting compressed data including a second compressed code obtained by compressing the reduced multi-valued image, position information of the character area, and representative color information of a unit character area in the character area;
First decompression means for decompressing the first compression code in the compressed data;
Second decompression means for decompressing the second compression code in the compressed data;
Enlarging means for enlarging the reduced multivalued image expanded by the second expanding means;
A synthesizing unit for synthesizing the binary image expanded by the first expansion stage and the multi-value image expanded by the expansion unit based on the position information and the representative color in the compressed data. Image decompression device. An image compression method for compressing an input multi-valued image,
A binarization step of binarizing the multi-valued image;
A first area specifying step of specifying a position of a character area in the binary image binarized in the binarizing step;
A second area specifying step of specifying the position of the unit character area in the character area specified in the first area specifying step;
A reducing step of reducing the multi-valued image,
A representative color extracting step of extracting a representative color of the unit character area based on the position information of the unit character area specified in the second area specifying step and the reduced multi-valued image;
A filling step of filling a unit character area in the character area of the reduced multi-valued image with its surrounding color based on the position information of the character area specified in the first area specifying step;
A first compression step of compressing the fill-in reduced multivalued image generated in the fill-in step;
A second compression step of compressing a partial binary image corresponding to the character area specified in the first area specification step;
The position information specified in the first and second area specifying steps, the first and second compressed codes generated in the first and second compression steps, and the unit character area extracted in the representative color extraction step An output step of outputting compressed data including the representative color information. An image decompression method for decompressing compressed data,
A first compression code for compressing a character area in a binary image obtained by binarizing a multi-valued image, and a unit character area in the character area of the reduced multi-valued image of the multi-valued image filled with surrounding colors. An input step of inputting compressed data including a second compressed code obtained by compressing the reduced multi-valued image, position information of the character area, and representative color information of a unit character area in the character area;
A first decompression step of decompressing the first compression code in the compressed data;
A second decompression step of decompressing the second compression code in the compressed data;
An enlarging step of enlarging the reduced multivalued image expanded in the second expanding step;
A synthesizing step of synthesizing the binary image decompressed in the first decompression stage and the multi-valued image expanded in the enlarging step based on the position information and the representative color in the compressed data. Image expansion method. A program for realizing image compression for compressing an input multi-valued image,
Program code for a binarization step of binarizing the multi-valued image;
A program code for a first area specifying step for specifying a position of a character area in the binary image binarized in the binarizing step;
A program code for a second area specifying step for specifying the position of the unit character area in the character area specified in the first area specifying step;
Program code for a reduction step of reducing the multi-valued image,
A program code for a representative color extracting step of extracting a representative color of the unit character area based on the position information of the unit character area specified in the second area specifying step and the reduced multi-valued image;
A program code for a filling process of filling a unit character region in the character region of the reduced multi-valued image with its surrounding color based on the position information of the character region specified in the first region specifying step;
A program code of a first compression step of compressing the fill-in reduced multi-value image generated in the padding step;
Program code for a second compression step of compressing a partial binary image corresponding to the character area specified in the first area specification step;
The position information specified in the first and second area specifying steps, the first and second compressed codes generated in the first and second compression steps, and the unit character area extracted in the representative color extraction step And a program code for an output step of outputting compressed data including the representative color information. A program for realizing image expansion for expanding compressed data,
A first compression code for compressing a character area in a binary image obtained by binarizing a multi-valued image, and a unit character area in the character area of the reduced multi-valued image of the multi-valued image filled with surrounding colors. A second compression code for compressing the reduced multi-valued image, position information of the character area, and program code of an input step of inputting compressed data including representative color information of a unit character area in the character area;
Program code for a first decompression step for decompressing the first compression code in the compressed data;
Program code for a second decompression step for decompressing the second compression code in the compressed data;
Program code for an enlarging step for enlarging the reduced multivalued image expanded in the second expanding step;
And a program code for a synthesizing step of synthesizing the binary image expanded in the first expansion stage and the multi-valued image expanded in the expansion step based on the position information and the representative color in the compressed data. A program characterized by the following.

Images