JP2003209704A - Image processing method, image processor, image forming device, image processing program, and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing method, an image processor, an image forming device, an image processing program, and a recording medium that can enhance the legibility of characters and symbols without changing a color tone of the characters and symbols and their background and provide an image easily read even by people with abnormity in color sensation. <P>SOLUTION: A legibility determination section 14 included in the image processor 1 of this invention extracts an area surrounded by edge parts as well as the edge parts from image data determined as the character edges as a character area and determines the legibility of the character area on the basis of a combination of colors of the character area and the surrounding area of the character area. A pattern comprising combinations of colors providing low legibility is prepared in advance for the determination of the legibility. Then a color correction section 15 applies correction processing for legibility improvement to image data of a character area whose legibility is determined low in the legibility determination section 14. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing method, an image processing apparatus, an image forming apparatus, an image processing program, and a recording medium for correcting image data forming a character or a symbol having low readability.

[0002]

2. Description of the Related Art Conventionally, an image processing apparatus has been developed which prevents turbidity of color characters and color misregistration in which colors are mixed with black characters. Japanese Patent Application Laid-Open No. 1-264847 discloses a color image processing device that detects the hue of an original image and the edge portion of a character to improve the quality of a reproduced image and controls the edge signal based on the hue signal. There is.

[0003]

[Patent Document 1] Japanese Patent Application Laid-Open No. 1-26484
The color image processing device described in Japanese Patent Publication No. 7 can prevent muddy color characters and deterioration of black character reproduction, but since the readability based on the combination of the character and the background color is not taken into consideration, the image quality of the reproduced image is Even if it is improved to some extent, there is a problem that readability is low. Readability of letters and symbols is
Influenced by the color of the background as well as the color of letters and symbols,
Further, it is also greatly affected by the difference in lightness of the colors of the characters, symbols and their background. Also, for people with color blindness (color blindness, color weakness), there are certain color combinations that are difficult to identify, so if the color combination of characters or symbols and the background corresponds to this particular combination, the reading It is desirable to correct the color of characters and symbols to provide easy images.

Further, the color image processing apparatus described in the above publication performs control processing involving changes in color components, such as enhancement of only necessary color components or deletion of unnecessary color components, on edge signals. Therefore, the color tone of the entire edge part including the outline of the character and the inside may change,
There may be a problem with the quality of the reproduced image.

An object of the present invention is to improve the readability of characters and symbols without changing the color tone of the characters and symbols and their background, and to provide an image that is easy to read even for people with color blindness. An image processing method, an image processing apparatus, an image forming apparatus, an image processing program, and a recording medium that can be performed.

[0006]

The present invention uses an extracting step of extracting image data of a character area from input image data, image data of the extracted character area and image data of an area surrounding the character area. Then, the image processing method comprises a readability determining step of determining the readability of the character area and a correction step of correcting the image data of the character area when the readability is determined to be low.

According to the present invention, a character or a symbol is distinguished as a character area and a background is distinguished as a peripheral area, and the readability is judged based on the combination of the two colors. Since the image data of the area is corrected, it is possible to improve the readability of characters and symbols that are difficult to read due to the combination with the background color.

According to the present invention, in the readability determination step, image data of the character area and the peripheral area are stored.
It is characterized in that the judgment is made by collating with a pattern formed by a combination of colors having a predetermined low readability.

According to the present invention, the readability is determined by collating the image data of the character area and the peripheral area with a pattern formed by a predetermined combination of colors having low readability, and therefore, processing such as calculation is not required. The readability can be easily determined. Moreover, since the processing is simple, the processing cost can be suppressed.

Further, the present invention is characterized in that the pattern is composed of a combination of colors which have low readability and are used for determination of color blindness.

According to the present invention, the pattern used for determining the readability is composed of a combination of colors having low readability used for determining the color blindness, and a combination of colors that is difficult to read by a person with color blindness is determined to have low readability. However, since the image data is corrected, it is possible to provide an image that is easy to read even for a person with color vision abnormality.

Further, the present invention is characterized in that in the readability determination step, the determination is performed using image data representing a hue of the character area and image data representing a hue of the peripheral area.

According to the present invention, the readability is determined by using the image data representing the hue of the character area and the image data representing the hue of the peripheral area. The colors of the background and the background can be accurately classified, and as a result, the readability can be accurately determined. Further, since the image data representing the hue can be easily calculated from the input image data, the processing cost can be suppressed.

Further, the present invention is characterized in that, in the readability determination step, the determination is performed by using image data representing the brightness of the character area and image data representing the brightness of the peripheral area.

According to the present invention, the readability is determined by further using the image data representing the brightness of the character area and the image data representing the brightness of the peripheral area. Therefore, the readability of the character, the symbol and the background is determined. Can be performed more accurately.

Further, the present invention is characterized in that the character area comprises an edge portion and an area surrounded by the edge portion, and in the correcting step, the image data of the edge portion is corrected.

According to the present invention, since the correction of the image data is performed on the image data of the edge portion which is the outline of the character or symbol, the color tone of the character region and the peripheral region, that is, the character or symbol and its background is changed. It is possible to improve readability without causing the problem.

Further, the present invention is characterized in that, in the correction step, the image data representing the brightness of the character area is corrected.

According to the present invention, since the image data is corrected for the image data representing the brightness of the character area, the color areas of the character area and the peripheral area, that is, the color of the character or symbol and the color of the background thereof are changed. It is possible to improve readability without causing the problem.

Further, according to the present invention, the extraction means for extracting the image data of the character area from the input image data, and the image data of the extracted character area and the image data of the peripheral area of the character area are used to extract the character area. Readability determining means for determining readability and color correcting means for performing color correction processing on input image data to obtain output image data are provided. When the color correcting means is determined to have low readability, the character area The image processing apparatus is characterized by correcting the image data of

According to the present invention, a character or symbol is distinguished as a character area and a background is distinguished as a peripheral area, and the readability is determined based on the combination of both colors. Since the image data of the area and the peripheral area is corrected, it is possible to improve the readability of a character or a symbol which is difficult to read by combining with the background color.

The present invention also provides an image forming apparatus for forming an image on a recording medium based on the image processing apparatus described above, an image input apparatus for reading input image data, and output image data output from the image processing apparatus. An image forming apparatus comprising: an output device.

According to the present invention, the reproduced image with improved readability of characters and symbols can be used for various purposes.

The present invention is also an image processing program for causing a computer to execute the method described above.

According to the present invention, it is possible to cause a computer to execute the image processing method.

The present invention is also a computer-readable recording medium in which the above-mentioned image processing program is recorded.

According to the present invention, it is possible to cause a computer to execute the image processing method.

[0028]

BEST MODE FOR CARRYING OUT THE INVENTION An image processing apparatus 1 and an image forming apparatus 2 of the present invention will be described. FIG. 1 is a configuration diagram of an image forming apparatus 2 including the image processing apparatus 1. The image forming apparatus 2 is, for example, a digital color copying machine, and is configured by connecting a color image input apparatus 20 and a color image output apparatus 21 to the image processing apparatus 1. The color image input device 20 is, for example, a CCD (Charge Coupled Device).
The image is reflected by a document, which is read by a CCD as an analog signal of RGB (R: red, G: green, B: blue) and is input to the image processing apparatus 1.
The color image output device 21 outputs the image data onto a recording medium such as paper and is not particularly limited, but examples thereof include an electrophotographic system and an inkjet system. As a result, the image processing apparatus 1 of the present invention can use a reproduced image in which the readability of characters and symbols is improved for various purposes.

The image processing apparatus 1 includes an A / D (analog / digital) converter 10, a shading corrector 11, an input tone corrector 12, an area separator 13, a readability determiner 14,
The color correction unit 15, the black generation / undercolor removal unit 16, the spatial filter processing unit 17, the output gradation correction unit 18, and the gradation reproduction processing unit 19 are included. Further, the area separation unit 13 has a character edge determination unit 13a, and the readability determination unit 14 is a character region determination unit 14a, the L ^* a ^* b ^* conversion unit 14b and the determination unit 1.
4c. Here, L ^* a ^* b ^* means CIE (CI
E: Commission Internationale de l'Eclairage: International Lighting Commission) Recommended color space coordinates, where L ^* represents lightness and a ^* and b ^* represent color coordinates. The character area determination unit 14a
Is an extraction unit, the determination unit 14c is a readability determination unit, and the color correction unit 15 is a color correction unit. The color correction means performs the correction processing of the present invention for correcting the image data when it is determined that the readability is low, and the CMY for the L ^* a ^* b ^* signals.
Color correction processing for converting into a signal is performed.

The analog signal read by the color image input device 20 is stored in the image processing device 1 through the A / D converter 1.
0, a shading correction unit 11, an input gradation correction unit 12,
The region separation unit 13, the readability determination unit 14, the color correction unit 15, the black generation / undercolor removal unit 16, the spatial filter processing unit 17, the output gradation correction unit 18, and the gradation reproduction processing unit 19 are sent in this order, and CMYK. It is output to the color image output device 21 as a digital color signal of (C: cyan, M: magenta, Y: yellow, K: black).

The A / D converter 10 converts an RGB analog signal into a digital signal. Shading correction unit 1
1 is the digital R sent from the A / D converter 10.
For the GB signal, the illumination system of the color image input device 20,
Processing for removing various distortions generated in the imaging system and the imaging system is performed. The input gradation correction unit 12 adjusts the color balance of the RGB reflectance signals from which various distortions have been removed by the shading correction unit 11, and at the same time, an image such as a density signal that is used in the image processing apparatus 1. The processing system converts the signal into a signal that is easy to handle.

The area separating section 13 separates the block of each pixel or a plurality of pixels in the input image data from the RGB signal into one of a character area, a halftone dot area and a photograph area. Then, the area separation unit 13 outputs an area identification signal indicating which area the pixel or block belongs to, based on the result of the separation, to the color correction unit 15, the black generation / undercolor removal unit 16,
The output signal is output to the spatial filter processing unit 17 and the tone reproduction processing unit 19, and the input signal from the input tone correction unit 12 is output to the readability determination unit 14 as it is.

A method of separating the regions in the region separating unit 13 is, for example, “Society of Image Electronics Engineers of Japan, Proceedings 90-06.
The method described in “-04” can be used.
In this method, the following determination is made within a block of M × N (M and N are natural numbers) pixels centered on the pixel of interest,
It is used as the area identification signal of the pixel of interest.

First, for the central 9 pixels in the block
The average value of the signal level (D_ave), And the average value
Is used to binarize each pixel in the block. Maximum pixel signal
No. level (D_max), Minimum pixel signal level (D_min) Also same
Sometimes ask. In the halftone dot area, the image signal changes in the small area.
Uses high movement and high density compared to the background
Then, the halftone dot area is identified. Specifically, the binarized data
From 0 in the main and sub-scanning directions
Number of changes to 1 K_HNumber of changes from 1 to 0 K_VSeeking
Threshold T_H, T_VK compared to_H, K_VWith
If the value exceeds the threshold value, the halftone dot area is set. Again background
In order to prevent erroneous determination that D_aveAnd D_maxand
D_minIs the threshold value B₁, B₂Compare with. D_max−
D_ave> B ₁, And D_ave-D_min> B₂, And K_H> T
_H, And K_V> T_VIf it is, it is judged as a halftone dot area.
It If the condition is other than the above, it is determined as a non-halftone dot region. This
The area determined as the non-dot area is the character edge determination unit 13
At a, it is determined whether or not it is a character edge. Here is the character
The other areas that were not judged to be
Is determined. In this way, the area separation unit 13 outputs the data.
The area to which the pixels separated by the area identification signal belong
L depending on the area^*a^*b^*Color to convert signal to CMY signal
Correction processing and black generation processing can be performed.

FIG. 2 is a block diagram of the character edge determination unit 13a. The character edge determination unit 13a includes an edge detection unit 131 that performs a predetermined process on each of the R, G, and B planes, and an OR circuit that calculates the logical sum of the output signals output from the edge detection unit 131 of each plane (logical sum). Circuit 132 and a threshold value setting unit 133. Edge detection unit 13
1 further includes a density difference determination unit 134. In the edge detection unit 131, a 3 × 3 pixel block centered on the target pixel is first formed using two line memories. Next, in the density difference determination unit 134, the density difference between the pixel of interest and the neighboring pixels located around it is calculated for each plane. If any one of the obtained density differences is larger than the threshold value, the output signal is'ON ', and if it is smaller, it is'OFF'. As the threshold value, THr is used for the R plane, THg is used for the G plane, and THb is used for the B plane. In addition, THr, TH
g and THb are values output from the threshold value setting unit 133. The output signal from the edge detection unit 131 of each plane is input to the OR circuit 132, and the logical sum is calculated. The value calculated by this logical sum is finally output as a character edge signal. That is, if an edge is detected in any one of R, G, and B planes, the pixel of interest is determined to be a character edge, the character edge signal becomes'ON ', and an edge is detected in all planes. If not, the character edge signal is'OFF '.

Then, this character edge signal is output to the readability determination section 14. The character edge signals for all pixels are stored in the memory of the image processing apparatus 1 or the like. The edge detection method is not limited to the above method, and an edge detection filter such as a Sobel filter may be used for detection.

The readability determination unit 14 extracts not only the edge portion but also the area surrounded by the edge portion as a character area from the pixel data for which the character edge is determined, and the character area and the peripheral area of this character area are extracted. Readability is determined based on the color combination of and. Readability is determined by converting the RGB signals into L ^* a ^* b ^* signals.

The color correction unit 15 performs a correction process for improving the readability of the image data of the character area determined by the readability determination unit 14 as having low readability. Then, in order to faithfully reproduce the colors, color correction processing for converting the L ^* a ^* b ^* signals into CMY signals is performed, and the CMY signals are output to the spatial filter processing unit 17. The conversion matrix is used as the process of converting the RGB signal into the L ^* a ^* b ^* signal performed by the readability determination unit 14 and the color correction process of converting the L ^* a ^* b ^* signal into the CMY signal performed by the color correction unit 15. And a method using a lookup table (LUT) are used. In the method using the conversion matrix, the conversion from RGB to L ^* a ^* b ^* and L ^* a ^* b ^* to CMY is performed by the formula (1).
And the matrix arithmetic expression of Expression (2).

[0039]

[Equation 1]

First, a combination of major CMY values is given to the image output device 21 to output a color patch and measure the color. Similarly, the color image measured by the color image input device 20 is read. Then, the CMY and RGB values corresponding to the L ^* a ^* b ^* value are obtained, and the constants a _{11 to} a ₃₄ and b ₁₁ to satisfy the relationship of L ^* a ^* b ^* , CMY and RGB.
b _{34 is obtained} by the method of least squares. In order to obtain more faithful color reproduction, it is sufficient to include higher-order terms higher than the second order.

[0041] As a method of using a LUT obtains the transformation matrix, and the value of L ^* a ^* b ^* to be output to the RGB signal input is outputted to L ^* a ^* b ^* CMY
And the values corresponding to RGB and L ^* a ^* b ^* , L ^* a ^* b ^* and CMY.
There is a method of creating a LUT by using a neural network and learning by using the neural network instead of describing the relationship of A by a conversion matrix. In addition,
The neural network is a model of the nervous system of a living organism, and is a method of optimizing a weight parameter of a pre-assumed coupling between units by an algorithm such as an error backpropagation learning method. Transformation matrix and LUT as above
Is set as table data corresponding to each area determined by area separation, and the table data is switched for each area by the area identification signal, whereby color correction processing can be performed and image quality can be improved.

Next, the black generation / undercolor removal unit 16 performs black generation and undercolor removal processing. The minimum value of CMY is M
As the IN signal, the black generation amount K and the undercolor removal amount UCR (Under Color Removal) are calculated from the MIN signal. The black generation amount K and the undercolor removal amount UCR are calculated by calculating the MIN signal for each pixel, and then using the table data of the black generation amount K and the undercolor removal amount UCR corresponding to the preset MIN signal, This is done by taking out the value corresponding to the calculated MIN signal. The table data is set for each area, and the setting is switched according to the area identification signal. In order to reproduce the black character well, the values of the black generation amount K and the undercolor removal amount UCR are increased in the character edge area so that the black character is reproduced in a state close to a single black color. Further, in the photographic area, in order to improve the reproducibility of the pattern, the black generation amount K is reduced,
Reproduces K signals with high contrast as little as possible. The black generation amount K and the undercolor removal amount UCR are calculated as in the following formula (3), and the undercolor removal amount UCR is calculated from the CMY signal.
Is subtracted to obtain a CMYK signal as a spatial filter processing unit 1
Output to 7. In addition, let the value calculated by Formula (3) be C ', M', Y ', and K'.

[0043]

[Equation 2]

The spatial filter processing section 17 performs a spatial filter processing by a digital filter on the image data of the CMYK signals output from the black generation / undercolor removal section 16 based on the area identification signal to obtain a spatial frequency characteristic. As a result of the improvement, processing is performed to prevent blurring of the output image and deterioration of graininess. For example, the character edge determination unit 13 of the area separation unit 13
The area separated into the character edges in a is subjected to the sharp enhancement processing of the spatial filter processing by the spatial filter processing unit 17 in order to improve the reproducibility of the black character or the color character,
Increase the amount of high frequency emphasis. At the same time, the gradation reproduction processing unit 19 performs binarization or multi-value conversion processing on a high-resolution screen suitable for reproducing high frequencies. on the other hand,
Regarding the area separated into halftone dots by the area separation unit 13,
The spatial filter processing unit 17 performs appropriate processing such as low-pass filter processing for removing the input halftone dot component.

The output gradation correction unit 18 performs an output gradation correction process for converting a signal such as a density signal into a halftone dot area ratio which is a characteristic value of the color image output device 21. Then, the gradation reproduction processing section 19 performs gradation reproduction processing (halftone generation) for finally separating the image into pixels and processing so that each gradation can be reproduced. For the areas separated into photographs by the area separating unit 13, binarization or multi-value conversion processing on the screen with emphasis on gradation reproducibility is performed.

The image data of the CMYK signals which have been subjected to the above-mentioned respective processes are once stored in the memory of the image processing apparatus 1 or the like, read out at a predetermined timing and input to the color image output apparatus 21. Each of the above processes is controlled by a control unit such as a CPU (Central Processing Unit) of the image processing apparatus 1.

Next, the readability determination unit 14 and the color correction unit 1
The details of the configuration and the procedure for performing the image processing of the present invention according to No. 5 will be described. FIG. 3 is a configuration diagram of the readability determination unit 14. As described above, the readability determination unit 14 includes the character area determination unit 14a, the L ^* a ^* b ^* conversion unit 14b, and the determination unit 1.
4c. When the character edge signal output from the character edge determination unit 13a is input to the character region determination unit 14a, the character region determination unit 14a extracts the character region.
Information about the extracted character area is output to the L ^* a ^* b ^* conversion unit 14b as a character area signal. The L ^* a ^* b ^* conversion unit 14b converts the RGB signal into an L ^* a ^* b ^* signal,
L ^* a ^* b of the character area extracted based on the character area signal
^{* The} signal is output to the determination unit 14c. In the determination unit 14c,
The readability of the character is determined based on the L ^* a ^* b ^* signal of the character area, and the determination result is output to the color correction unit 15 as a readability determination signal together with the character area signal.

The character area determination unit 14 will be described with reference to FIGS.
The extraction processing of the character area in a will be described. Figure 4
FIG. 7 is a diagram showing a process of extracting an edge portion that is the contour of a character area. The outer contour of the character area is extracted in FIG. 4A, and the inner contour of the character area is extracted in FIG. 4B. Based on the character edge signal output by the character edge determination unit 13a, the outer contour of the character area is extracted, and then the inner contour of the character area is extracted. Extract.

Specifically, first, the character area determination unit 14a
The image data is scanned from left to right by raster scanning from the upper line to the lower line, and the character edge pixels determined as the character edge by the character edge determination unit 13a are extracted. Among the character edge pixels searched first, a pixel having a pixel which has not been processed yet and which is not the character edge on the left is used as the starting point. In FIG. 4, the starting points are shown as squares filled with a mesh pattern. The character edge pixels in the vicinity are searched from this starting point. In order to search for a character edge pixel in the vicinity, when extracting the contour outside the character area, as shown on the left side of FIG. 4A, the search is performed clockwise from the pixel on the right side of the start point. . When searching for a neighborhood other than the start point, FIG.
As shown in the center diagram of FIG. 3, the search is performed from the character edge pixel searched last time, that is, the contour pixel that is the contour of the character area. However, the contour pixel itself does not fall within the search range, and the search is actually performed from the next pixel. Here, the pixel located at the center of the search is the target pixel. As a result, the outer contour can be extracted in the clockwise direction as shown on the right side of FIG. It should be noted that the search range of the contour pixel may be either the pixel range of 4 or 8 near the pixel of interest, but the pixel range of 8 is the search range in FIG.

The extraction processing of the contour inside the character area is performed in the same procedure. First, among the pixels inside the outer contour extracted as described above, the pixel having the first searched character edge pixel and the pixel which is not the character edge on the right is used as the start point. Then, the character edge pixels in the vicinity are searched from this starting point. To search for a character edge pixel in the vicinity, the pixel located at the lower right of the start point is searched clockwise as shown in the diagram on the left of FIG. When searching in the vicinity of the pixel of interest other than the starting point, as shown in the center diagram of FIG. 4B, the contour pixel searched in the clockwise direction is searched, but this contour pixel itself falls within the search range. Instead, the search is actually performed from the next pixel. by this,
The inner contour can be extracted in the counterclockwise direction as shown in the right diagram of FIG. In both the outside and inside extraction processing, the processing ends when the pixel of interest returns to the start point.

5 to 8 show the flow of contour extraction processing. FIG. 5 is a flowchart showing the contour extraction processing in the character area determination unit 14a. First, step a1
Then, the contour extraction processing is started. At step a2, the starting point is selected. At step a3, the pixel of interest is selected. The pixel of interest is selected according to the above-mentioned predetermined conditions. In step a4, it is judged whether or not the selected target pixel is the start point, and if the target pixel is the start point, step a
5. If the pixel of interest is not the starting point, go to step a.
Return to 3. At step a5, the contour extraction process is terminated.

FIG. 6 is a flowchart showing the starting point selection processing of step a2 in FIG. Step a
In 2, the starting point selection process starts. Step b
In 1, the image is raster-scanned. In step b2, it is determined whether or not the target pixel is a pixel determined to be a character edge. If the pixel is an edge-determined pixel, that is, a character edge pixel, the process proceeds to step b4. If the pixel is not edge-determined, the process proceeds to step b3. In step b3, it is determined whether or not all the pixels have been searched. If all the pixels have been searched, the process proceeds to step a5, and the processing ends because the starting point is not found. If all pixels have not been searched, the process returns to step b1 to continue selecting the starting point. If the target pixel is an edge-determined pixel, it is determined in step b4 whether or not the detection flag is set, that is, whether or not it has already been detected as a character area. If the detection flag is set, the process returns to step b1 to continue selecting the next start point, and if the detection flag is not set, the process proceeds to step b5. In step b5, the detection flag is set with the pixel for which the detection flag is not set as the starting point. In step b6, the found starting point is set as the pixel of interest. At step b7, the process returns to step a2 in the flowchart of FIG.

FIG. 7 is a flow chart showing the process of selecting the target pixel in step a3 in FIG. At step a3, the process of selecting the target pixel is started. At step c1, a detection flag of the target pixel is set. Step c
In 2, in order to find the next pixel of interest, a search for neighboring pixels is performed. It should be noted that whether the neighborhood pixels are 8 neighborhoods or 4 neighborhoods is designated in advance. Then, in step c3, the process returns to step a3 in the flowchart of FIG.

FIG. 8 is a flow chart showing the process of searching for neighboring pixels in step c2 in FIG. In step c2, the search processing for neighboring pixels is started. In step d1, it is determined whether or not the pixel of interest is the starting point. If it is the starting point, the process proceeds to step d2. If it is not the starting point, the process proceeds to step d3. In step d2, the search is started from the pixel on the right or lower right of the start point. It should be noted that when extracting the contour outside the character area, it starts from the pixel to the right of the start point, and when extracting the contour inside the character area, it starts from the pixel to the lower right of the start point. In step d3, the search is performed clockwise by 1 from the contour pixel selected by the previous search.
Start from the advanced pixel. In step d4, character edge pixels are searched clockwise. At step d5, the character edge pixel found first is selected as the pixel of interest.
In step d6, step c2 in the flowchart of FIG.
Return to.

After extracting the inner and outer contours of the character area, the entire character area is extracted by extracting the area between the outer and inner contours, that is, the area surrounded by the edge portion serving as the contour. Extract. The following filling algorithm is used to extract the area surrounded by the edge portion that becomes the contour. FIG. 9 is a diagram showing the filling processing of the area surrounded by the edge portion as the contour. First, as shown in FIG. 9, when pixels forming the outer and inner contours of the character area are extracted, the rectangular area circumscribing the outer contour pixels is raster-scanned. Then, when the character edge pixel is found, the area from the character edge pixel to the next character edge pixel is filled, and the filled area is extracted as the character area.

Next, the readability determination process in the determination unit 14c will be described. Readability is greatly affected by the difference in brightness between the characters and symbols and the background color, but it is not solely determined by the difference in brightness. Therefore, the readability can be determined as follows, for example.

To determine the readability, first, the hue of the pixel determined to be the character area by the character area determination unit 14a and the hue of the pixels included in the peripheral area are calculated. In the L ^* a ^* b ^* conversion unit 14b, the RGB signal is converted into C by using the above equation (1).
Convert to IE1976L ^* a ^* b ^* . FIG. 10 is a diagram showing a region to be determined when the determination unit 14c determines readability. As shown in the figure, a rectangular area circumscribing the character area is obtained, and the rectangular area enlarged by n pixels is set as the target area. Note that n is an integer. _The average L ^* a ^* b ^* values of the pixel of the character region L ^*, a _in ^*, and b _{in the} ^* average L of the pixels included in the peripheral region other than the character region in the target area
Let ^* a ^* b ^* values be _outside L ^* , _{outside a} ^* , _outside b ^* . From these values, the hue H and the saturation C ^* are calculated by the following equations (4) and (5). It should be noted that, _{in the} L ^*, _the a ^*, _the b ^*
Calculated from the hue and _the value of saturation H and _the C ^*
And then, L _out ^*, a _outside ^*, b _{^* See} the value of the calculated hue and saturation from an H _{extracellular} and C _out ^*.

[0058]

[Equation 3]

FIG. 11 shows the colors of the character area and the peripheral area.
It is a figure which shows the color space coordinates to classify. Figure 11 (a) is color
Space coordinate is L^*FIG. 11 (b) is a view seen from the axial direction.
Is the color space coordinate b^*It is the figure seen from the axial direction. Illustration color sky
Inter-coordinate and calculated H_WithinAnd C_Within ^*And H_OutsideAnd C
_Outside ^*Depending on, the character area and the surrounding area are blue, red,
Be in any of the yellow, green, white, gray, or black color categories
To determine. Hue H is hue angle H₃To H₀In the range of
And the saturation C^*Is C₀ ^*If above, red, hue H is the hue angle
H₀To H₁And the saturation C^*Is C₀ ^*If more
Yellow, hue H is hue angle H₁To H₂In the range of and
Degree C^*Is C₀ ^*If it is above, green, hue H is hue angle H₂To H₃
And the saturation C^*Is C₀ ^*If it is more than blue
It Saturation C^*Is C₀ ^*If it is less than, it is achromatic and
Lightness L^*Is 0 or more L₀ ^*Black if below, L₀ ^*Greater than
L₁ ^*If less, ash, L₁ ^*If greater than 100
White. This makes it possible to
That is, the color classification of letters and symbols and their backgrounds is targeted.
Can be performed accurately, and as a result, the readability can be accurately determined.
It can be carried out. In addition, input the image data that represents the hue.
Since it can be easily calculated from image data,
The cost can be reduced. Text area color category
R_Within, R for the color category of the surrounding area_OutsideAnd Of hue
H separating categories₀, H₁, H₂, H₃The angle is
The present invention is not limited to this, and may be changed as appropriate.
It is possible. The color categories of the character area and the surrounding area are
If judged, category R_WithinAnd category R_OutsidePair with
Readability is judged from the combination.

FIG. 12 is a diagram showing a pattern made up of a combination of colors having low readability. Is prepared a pattern as shown in FIG advance, any of the patterns, investigated whether the combination of the color of the category R _outside the color category R _in a peripheral area of the character region are true, the character region Readability is determined. If there is a corresponding one, it is determined that the readability of the character area is low, and if there is no corresponding one, it is determined that there is no problem in the readability of the character area. Then, the determination result indicating that the readability is low is output as the readability determination signal. As a result, the readability can be easily determined, and the processing is simple, so that the cost required for the image data correction processing of the present invention can be suppressed. If it is determined that the readability of the character area is low, the following processing is further performed.

First, the absolute value of the difference between L- _inside ^* and L- _outside ^* is calculated, and this is taken as the lightness difference ΔL ^* (ΔL ^* = | L- _out ^* -L
_Internal ^* |), compared to a pre-determined threshold ΔLt ^*. If the lightness difference is equal to or larger than the threshold value (ΔL ^* ≧ ΔLt ^* ), the lightness difference is sufficiently large that there is no problem in readability, and the readability determination signal as the determination result is changed. If the lightness difference is smaller than the threshold value (ΔL ^* <ΔLt ^* ), the readability is low,
The readability determination signal is not changed. As a result, the readability can be determined more accurately.

Next, the correction of the present invention in the color correction section 15
The processing will be described. In the color correction unit 15, the determination unit 14
Based on the readability determination signal output from c, the image data
Data correction. That is, it was determined that the readability was poor.
In case L_Within ^*And L_Outside ^*Change the value of
It

[0063] _the average value L of the character region ^* the average value L _out of the peripheral area ^* and comparing, L _out ^* is large (L _in ^* <L _out ^*) when the value of the brightness of the character region It is changed to L _{in the} ^* ', which is calculated by the following equation (6). L _{in the} ^* '= L _in ^{* -ΔLt *} ... (6)

[0064] _in L ^* 'if becomes 0 or less, the value of the brightness of the peripheral area value L _out ^* calculated by the following equation (7)' was changed to the value of the brightness of the character region 0 _{(in the} L ^* '=
Change to 0). _Outside L ^* ′ = _Outside L ^* + (0- _Inside L ^* ′) (7)

[0065] On the other hand, when L _in ^* is large (L _in ^*> L _out ^*) changes the value of the brightness of the character region, the L _in ^* 'calculated by the following equation (8). L _{in the} ^* '= L _in ^* + ^{ΔLt *} ... (8)

[0066] In the case _{where the} L ^* 'has become more than 100,
The value L calculated by the following formula (9) for the brightness value of the peripheral area
Change to _outside ^* ', and change the lightness value of the character area to 100 ( _inside L ^* '
= 100). _Outside L ^* '= _Outside L ^* -( _Inside L ^* '-100) (9) If it is determined that there is no problem in readability, the correction process of the present invention is not performed.

As described above, by enlarging the brightness difference between the pixels in the character area and the peripheral area,
That is, it is possible to improve the readability of a character or symbol that is difficult to read by combining the character or symbol with the background color without changing the color tone of the character or symbol.

The color correction section 15 further performs color correction processing for converting the L ^* a ^* b ^* signal into a CMY signal, and this CMY signal is processed.
The signal is output to the spatial filter processing unit 17. This processing is performed by using, for example, the above-mentioned formula (2).

Next, the correction processing of the present invention based on the combination of colors that is difficult for a person having color vision deficiency to identify will be described. FIG. 13A is a diagram showing a combination of colors that have low readability and are used for determining color blindness, and FIG. 13B is a diagram showing a combination of these combinations set as a pattern used in the correction processing of the present invention. Is. The first color blindness refers to a state in which it is difficult to distinguish red from its complementary color,
Color blindness is a state in which it is difficult to distinguish between green and its complementary color. The determination unit 14c, to one of the pattern, _the category R color character area and a color category R _outside the peripheral region
The readability of the character area is determined by checking whether or not the combination with and is applicable. If there is a corresponding one, it is determined that the readability of the character area is low, and if there is no corresponding one, it is determined that there is no problem in the readability of the character area. Then, the readability determination signal indicating that the readability is low, which is the determination result, is output to the color correction unit 1.
Output to 5.

Then, in the color correction section 15, the hue of the character area is rotated or the lightness value of the character area is converted,
Corrects the hue or brightness of the character area and corrects characters and symbols. In this correction, the hue can be rotated or the lightness value can be converted independently, or both the hue can be rotated and the lightness value can be converted. Rotation of hue is performed using the following formula. In addition,
hue angle for rotating the ^{θ, a * ', b *} ' after the rotation of the a ^*, b ^*
And

[0071]

[Equation 4]

The lightness value conversion of the pixel is performed in the same manner as in the above method, and when the hue rotation and the lightness value conversion are both performed, these may be simultaneously performed. The hue angle θ is shown in FIG.
A predetermined value can be set according to the pattern of (b). As a result, it is possible to provide an image that is easy to read even for a person with color vision deficiency.

Further, the correction processing of the present invention in the edge portion of the character area will be described. When the readability determination signal is output from the determination unit 14c to the color correction unit 15,
In addition to the readability determination result, a signal for determining whether the target pixel is a pixel in the edge portion of the character area or a pixel inside the character area, that is, a pixel surrounded by the edge portion is simultaneously output. That is, a total of three types are output, a signal indicating low readability and a character edge portion, a signal indicating low readability and inside a character area, and a signal indicating good readability. Then, the color correction unit 15 performs the correction processing of the present invention on the character edge portion having low readability, and does not perform the correction processing on the inside of the character area having low readability, or a calculation formula used for the correction processing is performed. Reduce the coefficient. Here, the portion to be corrected as the character edge portion may be limited to the contour pixel extracted by the character area determination unit 14. By performing the correction processing that emphasizes the brightness difference between the edge portion of the character area and the peripheral area in this way, the readability is improved without changing the color tone of the character area and the peripheral area, that is, the characters and symbols and their backgrounds. Can be improved.

FIG. 14 is a flow chart showing the image processing method of the present invention. At step e1, image processing is started. In step e2, the character area determination unit 14a
The pixels of the character area are extracted by. In step e3, the L ^* a ^* b ^* conversion unit 14 determines the readability.
At b, the RGB signals are converted into L ^* a ^* b ^* signals. In step e4, the determination unit 14c determines whether or not the readability of the character is low, using the pixel value of the character area and the pixel value of the peripheral area. The determination can be performed by comparing the pixel values of the character region and the peripheral region with a pattern formed of a predetermined combination of low-readable colors. This pattern may be composed of a combination of colors that have low readability and are used for determining color vision abnormality. The pixel value used for the determination may be hue,
It may be a brightness value. When it is determined that the readability is low, the process proceeds to step e5, and when it is determined that the readability is not low, the process proceeds to step e6. In step e5,
The value of the pixel in the character area is corrected based on the readability determination signal output by the determination unit 14c. This correction process can be performed for the brightness value of the character area,
It can also be performed for pixels in the edge portion of the character area.

The present invention can be realized by an image processing program for executing the image processing method. This program may be directly stored in the main memory of the image processing apparatus 1 that executes the image processing method, or may be stored in a recording medium that can be read by being inserted into a fixed disk device.

The directly stored program may be a process itself executed by a general-purpose computer such as a personal computer or a workstation, or a program loaded by using a communication network such as the Internet. May be. Further, the program may be a program generated after the program processing device performs a process such as decompressing the compressed data based on the loaded program. This allows the computer to execute the image processing method of the present invention.

The present invention is also realized by a recording medium recording the above image processing program. As a recording medium on which this program is recorded, for example, a tape system such as a magnetic tape and a cassette tape, an FD (Flexible D
magnetic disks such as isk) and HD (Hard Disk) and CD-ROM (Compact Disk Read Only Memor)
y), MO (Magneto Optical), MD (Mini Disk), DVD (Digital Versatile Disk), and other optical disk systems; IC (Integrated) including memory cards
Cards such as Circuits) and optical cards, mask ROM, EPROM (Erasable Programmable Read)
-Only Memory), EEPROM (Electrically Erasabl
A medium for fixedly storing the program can be used including a semiconductor memory such as an e Programmable Read-Only Memory) and a flash ROM. In any recording medium, the stored program may be configured to be accessed and executed by a computer, or the program read from the recording medium and loaded into the program storage area of the computer is executed. The configuration may be different. This allows the computer to execute the image processing method of the present invention.

[0078]

As described above, according to the present invention, it is possible to improve the readability of characters and symbols which are difficult to read due to the combination with the background color.

Further, according to the present invention, readability can be easily determined. Moreover, since the processing is simple,
The processing cost can be suppressed.

Further, according to the present invention, it is possible to provide an image that is easy to read even for a person having a color vision deficiency.

Further, according to the present invention, it is possible to accurately classify colors of characters, symbols and backgrounds, and as a result, it is possible to accurately determine readability. Further, since the image data representing the hue can be easily calculated from the input image data, the processing cost can be suppressed.

Further, according to the present invention, the readability of characters, symbols and the background can be determined more accurately.

Further, according to the present invention, readability can be improved without changing the color tone of the character or symbol and the background thereof.

Further, according to the present invention, a reproduced image in which the readability of characters and symbols is improved can be used for various purposes.

Further, according to the present invention, it is possible to cause a computer to execute the image processing method.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an image forming apparatus 2 including an image processing apparatus 1 according to an embodiment of the present invention.

FIG. 2 is a configuration diagram of a character edge determination unit 13a.

FIG. 3 is a configuration diagram of a readability determination unit 14.

FIG. 4 is a diagram illustrating a process of extracting an edge portion that is a contour of a character area.

FIG. 5 is a flowchart showing contour extraction processing in a character area determination unit 14a.

FIG. 6 is a flowchart showing a starting point selection process of step a2 in FIG.

FIG. 7 is a flowchart showing a target pixel selection process of step a3 in FIG.

8 is a flowchart showing a search process for neighboring pixels in step c2 in FIG. 7.

FIG. 9 is a diagram illustrating a filling process of a region surrounded by an edge portion that is a contour.

FIG. 10 is a diagram showing a region to be determined when the determination unit 14c determines readability.

FIG. 11 is a diagram showing color space coordinates for classifying colors of a character area and a peripheral area.

FIG. 12 is a diagram showing a pattern formed by a combination of colors having low readability.

FIG. 13A is a diagram showing a combination of colors having low readability used for determining a color vision abnormality, and FIG. 13B is a diagram showing a combination of these combinations set as a pattern used in the correction processing of the present invention. Is.

FIG. 14 is a flowchart showing an image processing method of the present invention.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Image processing apparatus 2 Image forming apparatus 10 A / D (analog / digital) conversion section 11 Shading correction section 12 Input tone correction section 13 Area separation section 13a Character edge determination section 14 Readability determination section 14a Character area determination section 14b L ^* a ^* b ^* conversion unit 14c determination unit 15 color correction unit 16 black generation / undercolor removal unit 17 spatial filter processing unit 18 output gradation correction unit 19 gradation reproduction processing unit 20 color image input device 21 color image output device 131 edge Detection unit 132 OR circuit (OR circuit) 133 Threshold value setting unit 134 Concentration difference determination unit

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Makoto Otsu             22-22 Nagaikecho, Abeno-ku, Osaka-shi, Osaka             Inside the company F term (reference) 5C077 MP05 MP07 MP08 PP27 PP28                       PP32 PP33 PP35 PP36 PP37                       PP55 PQ12 PQ22                 5C079 HB01 HB03 HB06 HB08 HB12                       LA06 LB02 MA01 MA11 NA06                 5L096 AA02 AA06 BA17 FA06 FA44                       GA40 GA41

Claims (11)

[Claims] 1. A readability of a character area is determined using an extraction step of extracting image data of a character area from input image data, and using the extracted image data of the character area and image data of an area surrounding the character area. And a correction step of correcting the image data of the character area when the readability is determined to be low. 2. The readability determination step is performed by comparing image data of the character area and the peripheral area with a pattern made up of a predetermined combination of colors having low readability. Image processing method. 3. The image processing method according to claim 2, wherein the pattern is composed of a combination of colors that are less readable and are used for determining color blindness. 4. The readability determination step makes a determination using image data representing a hue of the character area and image data representing a hue of the peripheral area. The image processing method described in 1. 5. The image processing method according to claim 4, wherein in the readability determination step, the determination is performed using image data representing the brightness of the character area and image data representing the brightness of the peripheral area. 6. The character area comprises an edge portion and an area surrounded by the edge portion, and in the correcting step, image data of the edge portion is corrected. The image processing method according to any one of the above. 7. The image processing method according to claim 1, wherein in the correction step, image data representing the brightness of the character area is corrected. 8. A readability of a character area is determined using extraction means for extracting image data of a character area from input image data, and image data of the extracted character area and image data of a peripheral area of the character area. Readability determining means and a color correcting means for applying color correction processing to input image data to obtain output image data, and when the color correcting means is determined to have low readability,
An image processing apparatus for correcting image data of the character area. 9. The image processing apparatus according to claim 8, an image input apparatus for reading input image data, and an image output for forming an image on a recording medium based on output image data output from the image processing apparatus. And an image forming apparatus. 10. An image processing program for causing a computer to execute the method according to claim 1. 11. A computer-readable recording medium on which the image processing program according to claim 10 is recorded.