JP2000306098A - Method and device for image processing and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To extract a white background area with high precision by using all information of R, G, and B by dividing color image data into multiple areas and deciding whether or not each area is a white background area according to the feature quantities of the respective divided areas. SOLUTION: An image input part 1 such as a scanner inputs a color image and generates planes (e.g. image with 256 gradations quantized with 8 bits) of R, G, and B signals. An area division part 2 divides each of the plane of R, G, and B into the same areas (character area, pattern area, etc.), according to the features of the image. A part 3 for binarization by areas performs dynamic binarization by the areas of the respective planes. Consequently, threshold values as many as the areas are found. A part 4 for feature quantity calculation by the areas finds the mean value of pixel density above a binarization threshold in an area to obtain a feature quantity. A white background decision part 5 decides whether or not an area is a white background area on the basis of the calculated mean of the density of RGB.

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 for determining a white background area in a color image, an image processing apparatus, and a recording medium on which an image processing program is recorded. Applied to the judgment process.

[0002]

2. Description of the Related Art In digital color copiers and the like,
In order to reproduce an image with high image quality, it is necessary to separate a character area and a picture area in the image with high accuracy, and various image area separation techniques have been proposed.

For example, in the apparatus described in Japanese Patent Application Laid-Open No. 10-108012, character / picture determination is performed by combining edge determination, halftone determination, and white background determination.

[0004]

In the above-described apparatus, since the extraction of the white background area is performed using one signal (for example, a green signal), a light yellow background or a light cyan background is used as the white background area. There is a problem of erroneous determination.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide an image processing method, apparatus, and recording medium for extracting a white background area with high accuracy by using all information of RGB. Is to provide.

[0006]

In order to achieve the above object, according to the present invention, the R, G, and B planes of a color image are divided into a plurality of regions, and an automatic threshold value is set within the divided regions. One feature is that binarization is performed to calculate a feature amount for each area, and when the feature amount calculated for each plane is equal to or greater than a predetermined threshold, it is determined to be a white background area.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be specifically described below with reference to the drawings. FIG. 1 shows the configuration of the embodiment of the present invention, and FIG. 2 shows the flow of the processing in FIG. In FIG. 1, 1 is an image input unit such as a scanner, 2 is a region dividing unit, 3 is a binarizing unit for each region, 4 is a feature amount calculating unit for each region, and 5 is a white background determining unit.

An image input unit 1 such as a scanner inputs a color image and creates an RGB signal plane (for example, an image of 256 gradations quantized by 8 bits). The area dividing unit 2 divides each plane of RGB into the same area (character area, picture area, etc.) based on the characteristics of the image. Alternatively, each plane of RGB may be divided into regions (blocks) of the same shape composed of m pixels × n pixels. FIG. 2 shows an example in which blocks are divided into blocks having the same shape.

The binarization unit 3 for each area performs dynamic binarization for each area of each plane. In the present invention, more accurate white background extraction can be realized by using dynamic binarization. The reason for using dynamic binarization will be described below.

That is, in an example of a character on a white background as shown in FIG. 6, as shown in FIG.
Pixels near (white) appear with high frequency. On the contrary,
In an example of a non-white background (non-white background) as shown in FIG. 8, halftone pixels appear with high frequency.

When a fixed threshold value for binarization is set and an average value of pixels equal to or higher than the threshold value is obtained, the average value of characters on a white background and the average value of a non-white background are almost the same. For this reason,
In some cases, characters on a white background cannot be separated from a non-white background. Therefore, even in such a case, in order to satisfactorily separate the white background from the non-white background, the present invention uses dynamic binarization.

As a method of dynamic binarization, for example, a discriminant analysis method is used. In this method, first, a histogram of densities is created, and a histogram of a sum of densities and a sum of squares is calculated. Next, the density as a candidate for the threshold is set to 255.
By subtracting 1 from (white) to 0 (black), the density sum on the white side, the density sum on the black side, the variance outside the class, and the variance within the class are calculated. The concentration at which the out-of-class dispersion / in-class dispersion is maximized is defined as the out-of-class / in-class dispersion peak. Finally, the density value is set as a threshold for dynamic binarization.

By performing the dynamic binarization as described above, threshold values for the number of regions are obtained. The feature amount calculation unit 4 for each area calculates an average density of pixels having a threshold value or more in the area. That is, for one area, RGB3
Find the average of the two. The white background determination unit 5 determines whether the area is a white background area or a non-white background area based on the average of the calculated RGB densities. As described later, the determination in the attention area can be finely adjusted using the determination result of the surrounding area.

A description will be given of the feature amount calculation unit for each area.
FIG. 3 shows a histogram of the density in a certain area in one plane. An average value of pixel densities equal to or higher than the binarization threshold value is obtained, and is used as a feature amount. This feature amount is calculated in all areas in all RGB planes.

FIG. 4 is a flowchart of the white area determination processing. The feature amount of the area calculated by the feature amount calculating unit 4 is M (Red) and M (Green) for each plane.
n) and M (Blue). Assuming that the thresholds for determining whether or not the background is white are thr, thg, and thb, respectively, M (Red), M (Green), and M (Blue)
If e) is larger than thr, thg, and thb, respectively (steps 101, 102, and 103), it is determined that the area is a white background area (step 104). Otherwise, it is determined as a non-white background area (step 105). This process is performed for all the regions, and the region determined to have a white background is output as a white background separated image (step 106).

FIG. 5 is a diagram for explaining a white area determination method using the determination result of the surrounding area. In the determination method of FIG. 4, there is an area where it is difficult to determine whether or not the area is a white background area. For example, it is an area composed of fine characters. Therefore, in order to prevent erroneous determination, the determination result of the surrounding area is used. For example, when the attention area is an area that is difficult to determine, referring to the determination result of the surrounding area,
If even one is determined to be a white background area, the attention area is determined to be a white background area.

In the above-described embodiment, an example in which a white background area is extracted from RGB image data has been described. However, the present invention is not limited to this, and an RGB signal is converted into a luminance signal Y and color difference signals I and Q. Alternatively, the image converted into the luminance signal may be divided into a plurality of regions, and whether or not the image is a white background region may be determined based on the density distribution of pixels in the region.

The present invention is not limited to the above-described embodiment, but can be realized by software. When the present invention is realized by software, FIG.
As shown in FIG. 1, a computer system including a CPU, a memory, a display device, a hard disk, a keyboard, a CD-ROM drive, a scanner, etc. is prepared.
On a computer-readable recording medium such as M, a program for realizing the white background area determination function of the present invention is recorded, and this program is installed in the system. A color image input from an image input unit such as a scanner is temporarily stored in a hard disk or the like. Then, when the program is started, the temporarily stored image data is read, the above-described determination processing is executed, and the result of the determined white background area is output.

[0019]

As described above, according to the present invention, a color image is divided into regions (blocks) and it is determined whether or not each divided region has a white background.
It is possible to accurately determine and extract a white background area.
Further, since each plane image is divided into a plurality of regions, and it is determined whether or not the region is a white background region based on the characteristic amount of each region (average of the density of pixels equal to or higher than a threshold value), A light cyan background is not erroneously determined as a white background. Furthermore, for an area in which it is difficult to determine whether or not the area is a white background area, the determination result of the surrounding area is used.
It is possible to extract a white background region with higher accuracy.

[Brief description of the drawings]

FIG. 1 shows a configuration of an embodiment of the present invention.

FIG. 2 shows a processing flow of the present invention.

FIG. 3 shows a histogram of density in a region.

FIG. 4 shows a processing flowchart of white background determination.

FIG. 5 is a diagram illustrating determination of a white background area using surrounding block information.

FIG. 6 shows an example of a character on a white background.

FIG. 7 shows a density histogram of a character on a white background.

FIG. 8 shows an example of a non-white background.

FIG. 9 shows a density histogram on a non-white background.

FIG. 10 shows a configuration example when the present invention is realized by software.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Image input part 2 Area division part 3 Binarization part for every area 4 Feature amount calculation part for every area 5 White background judgment part

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) H04N 1/40 103C 1/46 Z F term (Reference) 5B057 BA02 CA01 CA08 CA16 CB01 CB06 CB16 CC02 CE12 DA08 DB06 DC23 DC25 5C077 MM01 MP08 PP27 PP32 PP46 PP80 RR02 TT06 5C079 HB01 LA01 LA06 LA34 PA02 5L096 AA02 AA06 BA07 EA43 FA32 FA37 GA19 GA40 9A001 EE05 HH25 HH31 JJ35 KK31 KK42

Claims (7)

[Claims] 1. An image processing method comprising: dividing color image data into a plurality of regions; and determining whether each of the divided regions is a white background region based on a feature amount of each of the divided regions. . 2. The image processing method according to claim 1, wherein the feature amount is a density distribution in the area. 3. The image processing method according to claim 1, wherein a plurality of plane images are generated from the color image data, and each of the plane images is divided into a plurality of regions. 4. The image processing method according to claim 1, wherein the color image data is converted into luminance image data and then divided into a plurality of regions. 5. The image processing method according to claim 1, wherein when the area is a predetermined area, it is determined whether or not the area is a white background area based on a determination result of an area near the predetermined area. . 6. A means for inputting color image data, a means for dividing the color image data into a plurality of areas, a means for binarizing each of the areas, Means for determining whether each of the areas is a white background area based on a density distribution. 7. A computer realizes a function of dividing color image data into a plurality of areas and a function of determining whether or not each area is a white background area based on a feature amount of each of the divided areas. Readable recording medium on which a program for recording is recorded.