JP2008109386A - Image encoding apparatus and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problems that pixel values outside an extraction part are substantially changed and quality at the time of decoding declines in the case that a difference between the pixel value around the extraction part of resolution information and the substituting pixel value of the extraction part is large, and the effect of smoothing an edge is reduced and it is difficult to improve the encoding efficiency of irreversible encoding when the change of the pixel values outside the extraction part is reduced by a conventional technique. <P>SOLUTION: The generation of a substituting value after resolution information extraction is performed so as to smooth an edge by using an interpolation function. At the time, the correlation of the pixel values within a block is calculated for main scanning and sub scanning directions, interpolation by the interpolation function is performed for the direction of the low correlation, and the edge is smoothed by using a linear LPF for the direction of the high correlation. Thus, the encoding efficiency of the irreversible encoding to gradation information is improved while keeping the quality at the time of decoding. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a technique for efficiently encoding an image in which a character or line drawing and a natural image are mixed.

  Conventionally, as a method for encoding a multi-value image in Patent Document 1 or the like, after dividing a multi-value image into blocks that are units of orthogonal transformation, the mode value with the highest appearance probability in the block is converted to a character or line drawing. Extracted as color information or density information (hereinafter referred to as “extracted color”), run-length encoding, etc. for the extracted color and position information of the extraction unit (hereinafter referred to as resolution information together) There is known a coding method that performs lossy coding such as JPEG for a natural image (hereinafter referred to as gradation information) after character and line drawing information extraction.

At this time, with respect to the extraction unit, the pixel value is replaced by the average value replacement of the pixels in the block excluding the resolution information extraction unit, the previous value replacement, etc. By performing the filter processing with the LPF), the edge between the extraction unit and the non-extraction unit pixel is smoothed to improve the encoding efficiency of lossy encoding for gradation information.
JP-A-4-326669

  However, when the difference between the pixel value around the extraction unit and the replacement pixel value of the extraction unit is large, the above-described conventional technique significantly changes the pixel values other than the extraction unit, and the quality at the time of decoding is reduced. was there.

  Further, if the change of pixel values other than the extraction unit is reduced, the effect of smoothing the edge is reduced, and it is difficult to improve the encoding efficiency of lossy encoding.

  The present invention has been made in order to solve the above-described conventional problems, and an object thereof is to improve the encoding efficiency of lossy encoding for gradation information while maintaining the quality at the time of decoding.

  Then, the generation of the replacement value after extracting the resolution information is performed using an interpolation function so that the edge becomes smooth. At this time, the correlation between the pixel values in the block is calculated in the main scanning and sub-scanning directions, and interpolation is performed using an interpolation function in the direction where the correlation is low. The LPF is used to smooth the edges.

  In order to solve the above problems, according to the image coding apparatus of the present invention, an input image is blocked by a predetermined size, an extracted color determining means for determining an extracted color in the block, and the block of interest Means for detecting the position of a pixel corresponding to the extracted color; means for generating a replacement value from a pixel other than the extracted color of the target block; and replacing the pixel corresponding to the extracted color of the target block by the replacement value generating means. Means for replacing with a value is provided, and the block image after replacement, the extracted color, and the position information of the extracted color are encoded separately.

  It is possible to provide an encoding device with high gradation information encoding efficiency.

(First embodiment)
FIG. 1 shows a configuration diagram of an encoding apparatus according to an embodiment of the present invention. 101 is an input terminal for inputting a multi-valued image, 102 is a buffer for temporarily storing image data, and 103 is a data separating unit for separating resolution information and gradation information from an image. .

  A replacement value calculation unit 104 calculates a replacement value of a portion extracted as resolution information. Reference numeral 105 denotes a selector, reference numeral 106 denotes a lossless encoding unit for performing run-length encoding on the extracted color, and reference numeral 107 denotes a lossless encoding unit for lossless encoding of the position information of the extracted portion.

  Reference numeral 108 denotes an irreversible encoding unit for irreversibly encoding gradation information such as JPEG compression. Reference numeral 109 denotes a multiplexing unit, which packs each piece of code data so that it can be easily stored in a subsequent memory, and outputs it as code data. Reference numeral 110 denotes a code data output terminal.

  Next, the operation of the encoding apparatus in FIG. 1 will be described.

  The multivalued image input from the input terminal 101 is temporarily stored in the buffer 102. Then, the image data is sequentially read out in units of blocks and input to the data separation unit 103 and the replacement value calculation unit 104. The data separation unit 103 determines the most frequent color in the block and outputs it as the extracted color to the lossless encoding unit 106.

  At the same time, the data separation unit 103 extracts the position information of the resolution information and outputs it to the replacement value calculation unit 104, the lossless encoding unit 107, and the selector 105. Here, FIG. 2 shows an example of image data in which bitmap information that is position information of resolution information is superimposed. Bitmap information is 1-bit information, a flag of “1” is set for the pixel position extracted as resolution information, and “0” is set for other pixels.

  Thereafter, the replacement value calculation unit 104 calculates the replacement value of the pixel extracted as the resolution information based on the bitmap information indicating the pixel position of the resolution information, and outputs it to the selector 105.

  The selector 105 selects the image data and the replacement value based on the bitmap information indicating the pixel position of the resolution information, and outputs it to the lossy encoding unit 108 as gradation information. That is, when the bitmap is ‘0’, the image data is selected and output, and when the bitmap is ‘1’, the replacement color is selected and output.

  Then, the lossless encoding unit 106 performs run-length encoding on the extracted color and outputs the result to the multiplexing unit 109. In the lossless encoding unit 107, the bitmap information of the resolution information is encoded and output to the multiplexing unit 107. Further, the lossy encoding unit 108 encodes the gradation information and outputs it to the multiplexing unit 109.

  Thereafter, the multiplexing unit 109 packs each piece of code data so that it can be easily stored in a subsequent memory, and outputs it as code data from the output terminal 110.

  Next, FIG. 3 shows a configuration diagram of the replacement value calculation unit 104. Reference numeral 301 denotes an input terminal for image data, and the image data is input in units of blocks of 8 pixels × 8 pixels. Reference numeral 302 denotes a correlation value calculation unit for calculating correlation values in the main scanning direction (x direction) and the sub scanning direction (y direction), and 303 denotes a comparator. An interpolation unit for generating a replacement value from the 304 interpolation function, 305 is an LPF, and 306 is an output terminal for outputting the replacement value.

  Next, the operation of the replacement value calculation unit in FIG. 3 will be described.

  First, image data is input in block units from the input terminal 301 to the correlation value calculation unit 302 and the interpolation unit 304.

  As an example of the image data, FIG. 4 shows a diagram in which the image data and the bitmap information of FIG. 2 are superimposed.

  The pixels shown in gray in FIG. 4 are pixels around the extraction unit that indicate the outline of the portion extracted as the resolution information, and the squares shown by the thick dotted lines indicate the block of 8 pixels × 8 pixels. In addition, pixels indicated by diagonal lines are block peripheral pixels, which are pixels that exist at the boundary of the block and are included in the extraction unit.

  Then, the correlation value calculation unit 302 calculates the correlation value x in the main scanning direction and the correlation value y in the sub-scanning direction, and outputs them to the comparator 303.

  In FIG. 2, assuming that the pixel value at the position of (column x, line y) is P (x, y), the correlation value x and the correlation value y are calculated by the following equations.

ここで、主走査方向の相関が低いほど相関値ｘが大きく、副走査方向の相関が低いほど相関値ｙの値は大きくなる。

Here, the correlation value x increases as the correlation in the main scanning direction decreases, and the correlation value y increases as the correlation in the sub scanning direction decreases.

  The comparator 303 compares the pixel values of the correlation value x and the correlation value y. When the correlation value x is larger, “1” is used, and when the correlation value y is larger, “0” is used as the determination signal. Output to the interpolation unit 304 and the LPF 305.

  The interpolation unit 304 performs an interpolation calculation in the main scanning direction or the sub-scanning direction according to the determination signal, and outputs the interpolated image data to the LPF 305.

  Hereinafter, the interpolation calculation method will be described with reference to FIG. Interpolation is performed according to the following rules.

  Rule 1: A pixel other than an extraction pixel (hereinafter referred to as an extraction pixel) in a line or column is used as a control point (passing point) to complement the extraction unit.

  Rule 2: When a pixel at the boundary of a line or column (hereinafter referred to as a boundary pixel) is an extraction pixel, the extraction unit is complemented with the boundary pixel as a control point (passing point).

  Rule 3: If there is no pixel other than the extracted pixel in the line or column, the extraction unit is complemented with the average value of the pixels other than the extracted pixel.

  Specifically, the image shown in FIG. 2 is performed as follows.

  First, if the average value of pixels other than the extracted pixels is M, M is calculated by the following equation.

M = (P (6,0) + P (7,0) + P (0,1) + P (6,1) + P (7,1) + P (3,4) + P (0,7) + P (7,7) )) / 8 (Formula 1-3)
When the determination signal is “1”, interpolation in the main scanning direction is performed as follows.

  According to rule 2, the interpolation of line 0 is performed using pixel (0, 0) and pixel (6, 0) as control points (pass points). The interpolation value replaces the extraction unit with P (6,0).

  According to rule 1, the interpolation of line 1 is performed using pixel (0, 1) and pixel (6, 1) as control points (pass points). Interpolation is performed by connecting two points with a straight line.

  Interpolation of the lines 2, 3, 5, 6 is performed by replacing the average value M as an interpolation value according to rule 3.

  According to rule 2, the interpolation of the line 4 is performed using the pixel (3, 4), the pixel (6, 0), and the pixel (7, 4) as control points (passing points). The interpolated value replaces the extraction unit with P (3,4).

  According to rule 1, the interpolation of the line 7 is performed using the pixel (0, 7) and the pixel (7, 7) as control points (pass points). Interpolation is performed by connecting two points with a straight line.

  When the determination signal is “0”, interpolation in the sub-scanning direction is performed as follows.

  Interpolation of column 0 is performed according to rule 2 using pixel (0, 0) and pixel (0, 1) as control points (passing points). The interpolation value replaces the extraction unit with P (0,1). Further, according to rule 1, pixel (0, 1) and pixel (0, 7) are used as control points (pass points). Interpolation is performed by connecting two points with a straight line.

  Interpolation of the columns 1, 2, 4, and 5 is performed by replacing the average value M as an interpolation value according to rule 3.

  According to rule 2, the interpolation of the column 3 is performed using the pixel (3, 0), the pixel (3, 4), and the pixel (3, 7) as control points (passing points). The interpolated value replaces the extraction unit with P (3,4).

  According to rule 2, the interpolation of the column 6 is performed using the pixel (6, 1) and the pixel (6, 7) as control points (passing points). The interpolation value replaces the extraction unit with P (6,1).

  The interpolation of the column 7 is performed according to the rule 1 with the pixel (7, 0) and the pixel (7, 7) as control points (pass points). Interpolation is performed by connecting two points with a straight line. Thereafter, the LPF 305 performs one-dimensional filter processing in the main scanning direction or sub-scanning direction according to the determination signal, and outputs the processed image data to the output terminal 306. The filter processing is performed in the main scanning direction when the determination signal is “0”, and is performed in the sub-scanning direction when the determination signal is “1”.

  As described above, the generation of the replacement value after extraction of the resolution information is performed so that the edge is smooth using an interpolation function for the direction of low correlation in the main scanning direction or the sub-scanning direction, For directions with high correlation, smoothing edges using a one-dimensional LPF can improve the coding efficiency of lossy coding for gradation information while maintaining the quality at the time of decoding. .

  In this embodiment, the interpolation between the two control points (passing points) is performed with a straight line, but the interpolation may be performed using another interpolation curve such as a cosine function, a spline function, or a Bezier curve.

  Further, in this embodiment, when the input data has a plurality of color components such as RGB, the replacement value is calculated by the replacement value calculation unit 104 for each color. At this time, when the direction having a strong correlation is different for each color, the direction of interpolation and filtering may be different for each color.

(Other embodiments)
A storage medium storing software program codes for realizing the functions of the above-described embodiments is supplied to a system or apparatus, and a computer (or CPU or MPU) of the system or apparatus stores program codes stored in the storage medium. It goes without saying that the object of the present invention is also achieved by executing the reading.

  In this case, the program code itself read from the storage medium realizes the novel function of the present invention, and the storage medium storing the program code constitutes the present invention.

  As a storage medium for supplying the program code, for example, a flexible disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a CD-R, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used.

  In addition, the functions of the above-described embodiments are realized by executing the program code read by the computer, and the OS running on the computer is part of the actual processing based on the instruction of the program code. Alternatively, the functions of the above-described embodiment can be realized by performing all of them and performing the processing.

  Further, after the program code read from the storage medium is written to a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer, the function expansion is performed based on the instruction of the program code. The CPU of the board or the function expansion unit performs part or all of the actual processing, and the functions of the above-described embodiments can also be realized by the processing.

  The present invention also applies to a case where the program is distributed to a requester via a communication line such as personal computer communication from a storage medium in which a program code of software that realizes the functions of the above-described embodiments is recorded. Needless to say, it can be applied.

It is a block diagram which shows the structure of the encoding apparatus in embodiment of this invention. It is a figure which shows the bitmap information of the resolution information in embodiment of this invention. It is a block diagram which shows the structure of the replacement value calculation part in embodiment of this invention. FIG. 3 is a diagram in which image data and the bitmap information of FIG. 2 are superimposed.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 Image input terminal 102 Buffer 103 Data separation part 104 Replacement value calculation part 105 Selector 106 Lossless encoding part 107 Lossless encoding part 108 Lossless encoding part 109 Multiplexing part 110 Code data output terminal

Claims (6)

Means for blocking the input image into a predetermined size;
Extraction color determination means for determining the extraction color in the block;
Means for detecting the position of a pixel corresponding to the extracted color of the block of interest;
Means for generating a replacement value from pixels other than the extracted color of the block of interest;
Means for replacing a pixel corresponding to the extracted color of the block of interest with a replacement value by the replacement value generating means;
An image encoding device, wherein the block image after replacement, the extracted color, and the position information of the extracted color are encoded separately. Correlation value calculating means for calculating the horizontal and vertical correlations of the image,
Comparing means for comparing the correlation value in the horizontal direction and the correlation value in the vertical direction and outputting a comparison result;
Based on the comparison result, a replacement value generating means for calculating the replacement value using an interpolation function for a direction where the correlation is weak;
The image coding according to claim 1, further comprising: a one-dimensional low-pass filter that performs smoothing on the replaced block image based on the comparison result in a direction having a strong correlation. apparatus.   3. The image encoding apparatus according to claim 2, wherein the replacement value generation unit interpolates the extraction unit using a pixel other than the extraction pixel in the line or column as a control point (passing point).   When the terminal pixel of the line or column is an extraction pixel, the replacement value generation means sets the pixel closest to the terminal pixel as the terminal pixel, and sets the pixel other than the extraction pixel in the line or column as a control point ( The image encoding apparatus according to claim 2, wherein the extraction unit is interpolated as a passing point.   5. The replacement value generation means interpolates the extraction unit with a block average value of pixels other than the extraction pixels when there is no pixel other than the extraction pixels in the line or column. Image coding apparatus. Block the input image with a predetermined size;
An extraction color determination step for determining an extraction color in the block;
Detecting a position of a pixel corresponding to the extracted color of the block of interest;
Generating a replacement value from pixels other than the extracted color of the block of interest;
A step of replacing a pixel corresponding to the extracted color of the block of interest with a replacement value by the replacement value generation step;
An image encoding method, wherein the block image after replacement, the extracted color, and the position information of the extracted color are encoded separately.

Images