JP2002271628A - Color image processing method, color image processing apparatus, and recording medium - Google Patents

Abstract

(57) [Summary] A high quality color image is formed by eliminating the trade-off between the smoothing of structures such as halftone dots and line patterns and the retention of edges by characters, photographs, and the like. An image of an original color-printed on paper is input, a local adaptive edge holding smoothing process is performed on a digital color original image, and a halftone dot and a halftone image are held while retaining edges of characters and photographs of the image. Smooth structures such as line patterns.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a color image processing method, a color image processing apparatus, and a computer-readable recording medium on which a program for executing the color image processing method is recorded. The present invention relates to a method for efficiently smoothing a structure such as a halftone dot which is not always necessary for reproducing characters and photographs.

[0002]

2. Description of the Related Art When a color image is formed by a color copying machine or a printer, many color prints to be input are subjected to mechanical limitations, such as halftone dots and line patterns which are not included in the original input image data. The output image is configured by the structure.
As a result of the improvement in the resolution of the scanner and the image processing, the structure of the input image has been directly taken into the digital image processing apparatus as digital data. When the structure of the input image is simply smoothed, the edges of characters, photographs, and the like originally included in the image are also smoothed, and the image quality may be degraded.

On the other hand, a facsimile apparatus disclosed in Japanese Patent Application Laid-Open No. Hei 7-322060 switches the processing method according to the content of an image, judges continuous edges as edges included in a character area, and performs edge emphasis. The other edges are smoothed. Also,
For example, references M. Nagao, T. Matsuyama, "Edge Preserving
Smoothing "(Computer Graphics and Image Processin
g vol. 9, pp. 394 407, 1979) (hereinafter referred to as Reference 2).

In the image processing method disclosed in Japanese Patent Laid-Open No. 2000-22939, for example, all areas are temporarily smoothed, and thereafter, edge enhancement is selectively performed on the areas. In the image processing method disclosed in Japanese Patent Application Laid-Open No. 10-208038, it is necessary to change the range for averaging so that the image does not blur when the range for averaging includes edges of characters and photographs. In the processing method disclosed in Japanese Patent Application Laid-Open No. 6-68251, the size of the window function is enlarged by calculating the variance inside the window function.

[0005]

SUMMARY OF THE INVENTION Japanese Patent Application Laid-Open No. 7-32206
Both the methods disclosed in Japanese Patent Publication No. 0 and Reference 2 tend to enlarge small differences, and when considering image processing that requires threshold processing for edge strength, the dynamics of the edge strength of the entire image are considered. Range compression occurs and is disadvantageous. Further, for example, when an image obtained by taking in a printed material printed on both sides with a scanner or the like is processed, the edge of the show-through portion is also emphasized, and the appearance of the image is deteriorated. Further, the image processing method disclosed in JP-A-2000-22939 is disadvantageous in reproducing small kanji fonts and the like. Furthermore, in each case,
When processing is performed with a single window size, it is necessary to calculate the variance inside a plurality of windows per pixel for all pixels, which is disadvantageous in terms of the convergence time of the processing and the calculation time.

In the processing method disclosed in Japanese Patent Application Laid-Open No. Hei 10-208038, since a region for calculating a statistic as a feature does not have azimuth selectivity, it must be calculated in a sufficiently small window. The disadvantage is that the window is too small if it is necessary to smooth the structure of.
Further, in the processing method disclosed in Japanese Patent Application Laid-Open No. 6-68251, it is necessary to repeatedly calculate about 20 times of the entire image in order to determine the size of the window of the window function.
Since the shape of the window function is constant, there is a disadvantage in processing natural images and the like.

The present invention solves such disadvantages and eliminates the trade-off between smoothing structures such as halftone dots and line patterns and retaining edges by characters, photographs, and the like, thereby achieving high quality color images. An object of the present invention is to provide a color image processing method, a color image processing apparatus, and a computer-readable recording medium on which a program for executing the color image processing method can be formed.

[0008]

SUMMARY OF THE INVENTION A color image processing method according to the present invention provides a local window for a digital color original image obtained by digitally inputting an original color-printed on a paper surface using a digital scanner or the like. By changing the smoothing method in accordance with the value of the evaluation function in (1), the structure such as a halftone dot or a line pattern is smoothed while retaining the edges of characters and photographs of the image.

Further, by changing the smoothing method,
Prevents emphasis on noise edges.

Further, by converting color information into luminance information once and processing the luminance information, the calculation speed is improved and the misregistration between colors is less likely to occur.

Also, only when the statistic calculated inside a window of a predetermined size is larger than a predetermined threshold, a smaller window is set inside the window to calculate the statistic. For example, there is a method in which the size of the window is set to 7 × 7 pixels and the statistic is set to the standard deviation of the luminance. However, the size of the window may be changed according to the resolution of the input image. Further, another value such as variance may be used as the statistic. Further, only when the statistic is smaller than the threshold value, each color component of the target pixel is replaced with the average value of each color component inside the window.

By setting small windows in advance for a plurality of directions, the structure of halftone dots and the like is smoothed while retaining the edges of characters and the like. For example, there is a method of setting the small window in four directions of up, down, left, and right. However, this direction may be appropriately increased.

Each color component of the pixel of interest is replaced with the average of each color component within the small window in which the standard deviation of luminance is minimum among the small windows.

Further, when the minimum standard deviation of the luminance exceeds the luminance threshold, the pixel of interest is used as it is in the output image without replacement, thereby preventing excessive smoothing around the character.

The color image processing apparatus according to the present invention is characterized by smoothing a structure such as a halftone dot while maintaining the dynamic range of the edge of the original image by setting the threshold value. Since halftone dots and line patterns are repetitive putters, the standard deviation and variance within a window of a certain size are expected to be smaller than the standard deviation and variance of the edge part in a character or photo area. . Therefore, this threshold may be determined by a statistical method from the entire image.

Further, as described above, by using windows of a plurality of sizes in a nested manner, a high-speed convergence speed of processing is realized.

Further, the processing is terminated when the value of each pixel in the image is no longer changed, but may be used without repeatedly calculating by appropriately setting a threshold value.

[0018]

FIG. 1 is a block diagram showing the configuration of a color digital copying machine according to the present invention. As shown in the figure, the color digital copying machine has an image input unit 1, a display unit 2, an operation unit 3, and a printer unit 4. The color image input from the image input unit 1 is stored in the RAM 6. A series of processes for forming an image is executed by the CPU 5 reading out a program stored in the ROM 7. Further, the progress of the processing and the results of the processing are displayed on a display device 2 such as a CRT.
The user inputs and designates parameters necessary for processing from the operation unit 3 when necessary. Further, the intermediate data created during the execution of the processing is stored in the RAM 6 and read, corrected and written by the CPU 5 as necessary. The image generated as a result of the series of processing is used for subsequent image processing, or is read from the RAM 6 and output by the printer unit 4.

When the color image input by the image input unit 1 is processed by the CPU 5, the color image data 11 input by the image input unit 1 is subjected to local adaptive edge processing by the preprocessing 12 as shown in the processing flow of FIG. A holding smoothing process is performed, and thereafter, a normal copy machine process 13 such as color correction or show-through correction is performed.
Is performed to generate an output image.

The details of the local adaptive edge holding / smoothing processing performed in the preprocessing 12 will be described with reference to the flowchart of FIG. When image data 11 of digital data is input from the image input unit 1, a pixel of interest (i, j) is replaced with an average value inside a large window 21 having a predetermined size, for example, as shown in FIG. In order to check whether or not the target pixel (i,
j) each average value μ of RGB of W × W pixels centered on j)
A standard deviation σ (i, j) of (i, j) and the luminance inside the large window 21 is calculated (step S4). The calculated standard deviation σ (i, j) and a predetermined threshold T
H (step S5), and when the standard deviation σ (i, j) is smaller than the threshold value TH, it is considered that there is no edge of a character or the like inside the large window 21 and the large window 21
Each color component of the pixel of interest (i, j) is replaced with the average value μ (i, j) of each internal color component (step S6). If the standard deviation σ (i, j) of the luminance inside the large window 21 is larger than the threshold value TH, it is considered that there is an edge of a character or the like inside the large window 21, and the edge is held and smoothed inside the large window 21. (Step S7). When performing this edge-preserving smoothing, nine small windows 2 having orientation selectivity shown in FIG.
Calculate the standard deviation of each average value and luminance of each of the RGB inside 2, specify a small window in which the calculated standard deviation is the minimum, specify the standard deviation as s_σ (i, j) and the average value as s_μ (i , J). And the standard deviation s_σ (i,
j) is compared with the threshold value TH (step S8). When the standard deviation s_σ (i, j) is smaller than the threshold value TH, the average value s_μ of each color component inside the small window is obtained.
(I, j) replaces each color component of the pixel of interest (i, j) (step S9). When the minimum standard deviation among the nine small windows is larger than the threshold value TH,
The pixel value of the pixel of interest (i, j) is output as it is (step S10). When the standard deviation σ (i, j) of the luminance inside the large window 21 is smaller than the threshold value TH, and when the standard window s_σ (i, j) of the internal luminance is the smallest among the nine small windows 22, If the window is composed of the nine pixels nearest to the pixel of interest (i, j), the pixel of interest is not processed in the next and subsequent repeated calculations. This prevents excessive smoothing of the pixels in the smooth area. Then, the above-described repetitive calculation is performed until there is no change in the pixel value in the image (step S1).
1, S12, S4). In this processing, when importance is placed on the reproducibility of a thin line of about one pixel, the threshold value TH may be appropriately set so that the calculation is not repeated.

In this way, the image 23 composed of the halftone dots shown in FIG. 5A can be subjected to local adaptive edge holding and smoothing processing as shown in FIG.

In the above description, the case where the local adaptive edge holding / smoothing processing program of the preprocessing 12 is stored in the ROM 7 in advance, but this local adaptive edge holding / smoothing processing program is stored in an external storage medium. Alternatively, the CPU 5 may read from an external storage medium.

[0023]

As described above, the present invention smoothes the structure of halftone dots and line patterns resulting from the printing technique while retaining the edges of characters and photographs, thereby obtaining halftone dots and thousands of lines. Color information can be extracted without being disturbed by a structure such as a line pattern.

Further, by sequentially narrowing the processing range, it is possible to reduce the amount of calculation while extracting color information without being disturbed by the structure of halftone dots or line patterns.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a color digital copying machine according to the present invention.

FIG. 2 is a processing flowchart illustrating image processing.

FIG. 3 is a time chart showing a local adaptive edge holding smoothing process.

FIG. 4 is an explanatory diagram showing a window.

FIG. 5 is a display diagram showing images before and after a local adaptive edge holding smoothing process.

[Explanation of symbols]

1; image input unit, 2; display unit, 3; operation unit, 4; printer unit, 5; CPU, 6; RAM, 7; ROM, 21; large window, 22;

Continued on front page F-term (reference) 5B057 CA01 CA12 CA16 CB12 CB16 CC02 CE02 CE05 DB02 DC16 DC23 DC36 5C077 LL05 LL19 MP02 MP07 MP08 NN02 PP02 PP27 PP28 PP46 PP68 TT06 5L096 BA07 DA01 FA06 FA33 FA43 FA44 FA45 FA45

Claims (13)

[Claims] 1. A smoothing method for a digital color original image obtained by digitally inputting an original color-printed on a paper surface using a scanner or the like is changed by an evaluation function value in a local window. A color image processing method characterized in that, while retaining edges of contours of photographs, characters, and the like, halftone dots, line patterns, and the like that constitute them are smoothed. 2. The color image processing method according to claim 1, wherein the evaluation function is determined based on a statistic calculated inside a window. 3. The color image processing method according to claim 2, wherein a statistic inside a large window centered on a pixel of interest has an edge due to an outline of a photograph, a character, or the like inside the window. And calculating a statistic using a smaller window inside the window only when it is determined that 4. The color image processing method according to claim 3, wherein the small window including the target pixel is applied to several directions around the pixel of interest. Processing method. 5. The color image processing method according to claim 3, wherein the statistic is calculated and evaluated using windows of different sizes, so that a halftone dot is not compressed in the dynamic range of edge strength inside the image. A color image processing method characterized by smoothing a structure such as a line pattern or a line pattern. 6. The color image processing method according to claim 3, wherein the statistic is calculated and evaluated using windows of different sizes, thereby performing edge-preserving smoothing without repeatedly calculating the entire image. A color image processing method comprising: 7. A smoothing method for a digital color original image obtained by digitally inputting an original color-printed on a paper surface using a scanner or the like is changed by an evaluation function value in a local window. Thus, a color image processing apparatus is characterized in that halftone dots, line patterns, and the like constituting the contours of photographs and characters are smoothed while retaining the edges. 8. The color image processing apparatus according to claim 7, wherein the evaluation function is determined based on a statistic calculated inside the window. 9. The color image processing apparatus according to claim 8, wherein an edge due to a contour of a photograph, a character, or the like is present in the inside of the window based on a statistic inside a large window around a pixel of interest. A color image processing apparatus that calculates a statistic using a smaller window inside the window only when it is determined that 10. The color image processing apparatus according to claim 9, wherein the small window including the target pixel is applied to several directions around the pixel of interest. Processing equipment. 11. The color image processing apparatus according to claim 9, wherein the statistic is calculated and evaluated using windows of different sizes, so that halftone dots can be obtained without compressing the dynamic range of the edge strength inside the image. A color image processing apparatus characterized by smoothing structures such as lines and line patterns. 12. The color image processing apparatus according to claim 9, wherein the statistic is calculated and evaluated using windows of different sizes, thereby performing edge-preserving smoothing without performing repeated calculations on the entire image. A color image processing apparatus, characterized in that: 13. A recording medium on which a processing program of the color image processing method according to claim 1 is recorded.