JP2003228691A - Slip read device and program for processing slip read - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve the efficiency of the specific processing of a picture for recognizing characters by removing a dropout color from a slip picture. <P>SOLUTION: A slip read device and a program consider the difference of density distribution in histograms 12, 14, and 16 of R, G, and B plane multilevel pictures 11, 13, and 15 of a read slip. The read slip is provided with a red frame 17 of a dropout color, and characters or numerics are entered into the frame by a user in black. As for an automatic determination threshold indicating the so called distribution characteristics of each histogram, an R plane is '187', and G and B planes are both '152'. Then, the R plane whose value is the maximum and whose difference with a plane which is the largest next to this is not less than a prescribed value is selected as a dropout plane, and the character recognition processing of the picture is executed. Also, the dropout plane is specified on the basis of the magnitude of the black pixel frequency in the AND synthetic picture or XOR picture of a binary picture of each plane and a Y binary picture. <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 a device for reading a form (for example, a deposit / withdrawal slip used in a bank) including a dropout collar portion and a user entry portion related thereto.
And a program for reading the form.

[0002] In particular, a plain image or the like in which the red character / number entry frame portion (dropout color portion) is erased from the read image of the form is automatically selected, and the R and R of the read image are selected. For example, an AND operation is performed on each of the G and B plane binary images and the luminance signal binary image to generate a composite image in which the entry frame portion is erased, and based on these selected images and the generated images, a user entry is performed. It recognizes letters and numbers.

[0003]

2. Description of the Related Art Conventionally, an optical character reading apparatus of a type in which a dropout color portion such as a character frame on a form is removed by AND operation of R, G, B plane binary images of the form is disclosed in, for example, Japanese Patent Laid-Open No. Hei 10 (1999) -31977. No. 6-243290.

In the case of this optical character reader, R, G,
By executing an AND operation for each bit of each plane binary image of B, the dropout color portion of the read form is set as a background pixel (white pixel), and its character / number portion is set as an information pixel (black pixel). Generating a composite image.

[0005]

In the case of such an optical character reading method, since it is premised that the AND operation of each plane binary image is performed, a single dropout color (for example, red color) is used. There was a problem that the processing efficiency for the form was poor.

In a plain image of a single dropout color, this is the character frame part (dropout color part).
Are all removed, and there is no particular advantage due to the AND operation of the image and another color plane image.

Therefore, according to the present invention, it is judged whether or not there is a single plane in each plane of the form image, which can be said to be equivalent to the removal of all dropout color portions. By selecting a plain image and using it for subsequent character / number recognition processing, it is intended to improve the efficiency of the processing for specifying the character recognition image with the dropout color removed from the form image.

In the case where such a single plane does not exist, when a composite image of binary images of each plane is generated, a luminance signal binary image (Y signal binary image) of the reading form is generated.
It is intended to increase the resolution of this composite image by also using.

[0009]

The present invention solves this problem as follows. (1) In a form reading device for optically reading a form including a dropout color part and a user entry part related thereto in a pixel unit, a density distribution calculating means for obtaining a pixel density distribution for each plane of the read image. Plane selection means for determining a characteristic value of the pixel density distribution and specifying a dropout plane based on the characteristic value. (2) In a form reading device which optically reads a form including a dropout color part and a user entry part related thereto, in a form reading device, binarizing means for binarizing each plane of the read image, And a plane selecting unit that determines a black pixel frequency or a white pixel frequency corresponding to the plane binary image and specifies a dropout plane based on the pixel frequency. (3) In a form reading device that optically reads a form including a dropout color part and a user-written part related thereto in pixel units, a plane binary image and a luminance signal binary image of the read image are generated. Image generating means for generating the first combined binary image by calculating the luminance signal binary image and one of the plane binary images, and then generating the first combined binary image and the first combined binary image. A second combined binary image is generated by calculation processing with the other one of the plane binary images, and a third combined binary image is calculated with the remaining one of the plane binary images. And an image synthesizing means for generating a synthetic binary image.

According to the present invention, as described in (1) and (2) above, the characteristic value of the pixel density distribution in each plane of the read image of the target form (for example, an automatic discrimination threshold value described later) and the two planes of each plane. The dropout plane is specified based on the black pixel frequency or the white pixel frequency corresponding to the value image, that is, without performing a combining process such as an AND operation between the plane binary images. As a result, the generation process of the form image in which the dropout color is removed is efficiently executed.

Since the sum of the black pixel frequency and the white pixel frequency in the plane binary image is 100%, the conditions for specifying the dropout plane are equivalent regardless of which frequency is used. For example, the step (s
The conditional expressions for the black pixel frequency in 12) and (s15) also become the conditional expression for the white pixel frequency by substituting "black pixel frequency = 100-white pixel frequency".

Further, by using the black pixel frequency or the white pixel frequency of the composite binary image (having a higher resolution than the plane binary image itself) of the luminance signal binary image of the read image and the plane binary image, Improves the accuracy of identifying the dropout plane.

Further, when the dropout planes of (1) and (2) cannot be specified as in the above (3), when generating a composite binary image of each plane, the luminance signal binary of the form is generated. It also performs arithmetic synthesis with images.
This increases the resolution of the composite binary image itself.

Further, in the above (1) to (3), the threshold for binarization and the local binarization method described later (FIGS. 17 to 19).
By adjusting the background separation threshold and target range, etc., and appropriately converting the faint portions of the characters and numbers entered by the user into black pixels, the accuracy of the dropout plane can be improved and the final output image can be obtained. The letters and numbers in are so-called thicker to improve their recognition accuracy.

The present invention is intended for a form reading apparatus having such characteristics, and further efficiently executes a process of generating a form image in a form in which dropout color is removed, and the form image is highly accurate. A form reading program for making things available is also targeted.

[0016]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of the present invention will be described with reference to FIGS.

In the following description, for the sake of convenience of explanation, in principle, a red dropout color form is targeted, and
The size of the form is about A5 size, and the total number of pixels of the read image is 1,557,790 pixels.

FIG. 1 is an explanatory view showing the outline (part 1) of the dropout plane detection method.

In FIG. 1, 11 is a red plane when the form is read by a two-dimensional color CCD, 12 is a histogram of this red plane multivalued image, and 13 is the form 2.
Green plane when read by a three-dimensional color CCD, 14
Is the histogram of this green plane multi-valued image, and 15 is
A blue plane when the form is read by a two-dimensional color CCD, 16 is a histogram of this blue plane multivalued image,
Reference numerals 17 denote red frames (dropout portions) on the form, respectively.

The horizontal axis of each of the histograms 12, 14 and 16 is the pixel density (brightness) showing 256 gradations from the black level of "0" to the white level of "255",
The vertical axis corresponds to the number of pixels. The black frame of each plain image is an image component of the table portion on which the form is placed when the form is read.

The two-dimensional color CCD for reading the form is formed in a checkered pattern composed of red, green and blue photoelectric conversion elements (see FIG. 15).

The detection method of FIG. 1 focuses on the difference in density distribution in the histograms of the red, green, and blue plane images.

The form to be processed has, for example, a white background portion and a red frame 17, and the user writes black characters and numerical values in the red frame.

Red, green, blue (R,
The histograms of the planes 11, 13 and 15 of (G, B) are as follows: In the case of the red plane, there are two peak portions and the density distribution of the automatic discrimination threshold value is "187"; In both cases, there are three peak portions and the density distribution has an automatic discrimination threshold value of “152”.

Here, the dropout type of the target form is automatically determined by specifying the plane where the automatic determination threshold of the histogram is the maximum. In the case shown, it is determined that the form is a red dropout color form. It should be noted that the method itself for obtaining the automatic discrimination threshold value of the histogram is publicly known (FIG. 16).
reference).

Automatic discrimination threshold (= 18) of the red plane 11
7) is the automatic discrimination threshold (= 1 for green and blue planes 13 and 15)
52), that is, histogram 1 for each plane
The peak states of 2, 14 and 16 are different because the reflectance of light with respect to a reddish portion such as the reddish frame 17 on the form is larger in the red wavelength than in the green wavelength and the blue wavelength.

The read output (density) of the form with respect to the red frame 17 is naturally larger in the red photoelectric conversion element than in the other green and blue photoelectric conversion elements. The read output of the red photoelectric conversion element is substantially equal to that of the background portion (white portion) of the form.

Therefore, red pixels (such as a red frame) drop out in the red plane of the read image of the form, and the density distribution of the red pixels in the histogram of the plane is, so to speak, that of the background pixels (white pixels). It becomes a merged form. Such a merge phenomenon does not occur in the green and blue planes where the red frame remains in the read image.

As a result, the automatic discrimination threshold, which is a measure of the density distribution of each plane, becomes larger in the red plane than in the other green and blue planes.

Of course, in the case of a blue dropout color form such as a withdrawal form, the density distribution of the blue frame is merged with the density distribution of the background pixels (white pixels) in the histogram of the blue plane.

In this way, the dropout color type of the form is automatically determined based on the histogram of each plane of the form image. That is, it is determined that the read form in the case of FIG. 1 is a red dropout color. The value used for this automatic determination may be any value corresponding to the contents of each plane histogram, and may be, for example, the average value of the pixel density distribution.

FIG. 2 is an explanatory diagram showing the overall configuration corresponding to the detection method of FIG.

In FIG. 2, reference numeral 21 is a small-capacity two-dimensional color CCD of 400,000 pixels for optically reading an A5 size form, and 22 is a pixel shift control for the two-dimensional color CCD ( For example, a form image with a total of 3.6 million pixels is obtained based on nine times of pixel shift processing: FIG.
5) Pixel shift means, 23 is a picked-up image synthesizing means for synthesizing, for example, each of the nine images obtained by this pixel shift processing, 24 is a shading correction means for uniformizing the shade of this synthesized image, and 25 is The R, G, B plane dividing means 26 to 28 for dividing the read image after the shading correction into red, green and blue plane images are red, green and blue held in a memory or a work area (not shown). , 29 are Y plane image images, 29 are Y plane image images (luminance signal image), and 30 are held in a memory or a work area (not shown). The Y signal image, 31 is a binary Y obtained by binarizing the Y signal image by the local binarization method (see FIGS. 17 to 19).
The local binarization unit for creating an image, 32 is a Y signal binary image held in a memory or a work area (not shown), 33-
Reference numeral 35 is a histogram calculation unit that executes a histogram calculation of each red, blue, and green plane image (R image, G image, B image), and 36 to 38 are threshold values for binarization from the histogram of each plane. Automatic discriminating threshold method processing unit to be obtained,
39 determines whether a single dropout color type can be specified based on the threshold value of each plane (see step (s6) in FIG. 14), and if it can specify, a plane image determination that selects that plane.・ Selector, 40
Is a selected image binarization unit that binarizes each pixel signal of this selected plane with the automatic discrimination threshold, 41 is a selected plane binary image held in a memory or a work area (not shown),
42 is the Y signal binary image 32 and the selected plane binary image 4
AND operation synthesis unit for executing AND operation with 1, 43
Is a composite image output unit for outputting this AND composite image, 4
Reference numeral 4 denotes an integrated block composed of the constituent elements 33 to 39, respectively.

The local binarization unit 31 and the selected image binarization unit 40 each have a character blurring processing function described later (1.3 times the automatic discrimination threshold value in FIG. 3, FIG. 1).
See 9 and 20).

3 to 5 are explanatory views showing the outline (part 2) of the dropout plane detection method. Figure 3 is R
In the case of the plane, FIG. 4 shows the case of the G plane, and FIG. 5 shows the case of the B plane.

3 to 5 and FIGS. 7 to 9 described later.
In consideration of the display size of the handwritten character portion of each plane on the drawing, only the upper left ¼ portion of the entire plane is shown.

In FIGS. 3 to 5, 17 is a red frame similar to that in FIG. 1, 51 is a local binarization method for the luminance signals obtained from the red, blue and green plane images of the reading form ( FIG. 17-
19)), a Y signal binary image 52 is a first red plane binary image 52 obtained by binarizing the red plane of the reading form with its automatic discrimination threshold (= 123), and 53 is a reading form. The second red plane binary image 54 obtained by binarizing the red plane of the image with the blur correction threshold value (= 159: 1.3 times the automatic determination threshold value) is the Y signal binary image 51 and the first red plane binary image 54. An AND composite image with the red plane binary image 52 (black pixel frequency = 11.823%), 55 is an AND composite image of the Y signal binary image 51 and the second red plane binary image 53 (black pixel frequency = 12.030%), 62 is a first green plane binary image obtained by binarizing the green plane of the reading form with the automatic discrimination threshold value (= 103), and 63 is a blurred green plane of the reading form. Two correction thresholds (= 133: 1.3 times the automatic discrimination threshold) The second green plane binary image turned into asking, 64, Y signal AND the composite image of the binary image 51 and the first green plane binary image 62 (black pixel frequency = 13.
653%) and 65 are AND composite images of the Y signal binary image 51 and the second green plane binary image 63 (black pixel frequency =
15.495%), 72 is the first blue plane binary image obtained by binarizing the blue plane of the reading form with its automatic discrimination threshold (= 111), and 73 is the blur correction of the blue plane of the reading form. Threshold (= 144: 1.3 of automatic discrimination threshold)
Second blue plane binary image obtained by binarization with
4 is an AND composite image of the Y signal binary image 51 and the first blue plane binary image 72 (black pixel frequency = 13.787).
, 75 is an AND composite image of the Y signal binary image 51 and the second blue plane binary image 73 (black pixel frequency = 16.%).
273%), respectively. The black level is set to the logical value "1".

3 to 5, the black pixel frequency in each AND composite image of the binary image of each plane of the read form and the Y signal binary image (luminance signal binary image) of each plane. It focuses on the difference between. These AND combined images are also images in which the noise part of the binary image of each plane before combining is removed. That is, the resolution is higher than that of the plain image before composition.

Here, AND composite image 5 of the red plane
Reference numerals 4 and 55 are binary images in which the dropout color (red frame 17 and the like) of the read form has almost disappeared and only the characters and guide information are included.

On the other hand, the AND composite image 64 of the green plane,
The AND composite images 74 and 75 of 65 and the blue plane are
The image has a shape including the dropout color (red frame 17 and the like) of the read form.

The frequency of black pixels in each of the red, blue, and green plane binary images is increased by increasing the binarization threshold value as shown in the figure. This is because the read density of the pixels corresponding to the characters written on the form, the thin line portion such as the red frame or the blurred portion is small, and the pixel is processed as the white level when binarization by the low threshold, that is, Because it disappears.

AND composite images 54, 55 of each plane,
The black pixel frequencies at 64, 65, 74, and 75 are (1) in the case of the red plane of FIG. 3, 11.823% (automatic determination threshold: 123), 12.030% (blurring correction threshold: 159 = 123
× 1.3) (2) In the case of the green plane shown in FIG. 4, 13.653% (automatic determination threshold: 103), 15.495% (blurring correction threshold: 133 = 103)
× 1.3) (3) In the case of the blue plane of FIG. 5, ・ 13.787% (automatic determination threshold: 111) ・ 16.273% (blurring correction threshold: 144 = 111)
× 1.3). It should be noted that these numerical values are values in the entire image of each plane (not only the image of the ¼ portion shown in the drawing).

By obtaining and comparing the black pixel frequencies in the AND composite image of the respective red, green, and blue planes, the plane with the lowest frequency (the red plane in the case of the illustrated form) is dropped out. It can be specified as a plane.

The number of pixels corresponding to the black pixel frequency difference between the AND composite images of these planes is (1) in the case of the red plane and the green plane in the above A5 size read form with 1,557,790 pixels in total: 28,507 pixels (1.83% of the whole: at the automatic discrimination threshold) 53,978 pixels (3.465% of the whole: at the threshold for blur correction) (2) In the case of the red plane and the blue plane,・ 30,595 pixels (1.964% of the whole: when automatic discrimination threshold) ・ 66,097 pixels (4.243% of the whole: when threshold for blur correction) (3) In the case of green plane and blue plane , 2,088 pixels (0.134% of the whole: when the automatic determination threshold value), 12,119 pixels (0.778% of the whole, when the blur correction threshold value).

By using the blur correction threshold value larger than the automatic discrimination threshold value in this way, the AND of the red plane is performed.
The black pixel frequency difference between the composite image and other AND composite images is enlarged. Therefore, the dropout plane selection process can be performed more reliably.

The dropout plane may be specified based on the black pixel frequency difference of each plane image itself instead of the black pixel frequency difference of the AND combined image.

FIG. 6 is an explanatory diagram showing the overall structure corresponding to the detection method of FIGS.

In FIG. 6, 21 to 38 are constituent elements having the same contents as the corresponding numbers in FIG. 2, and 81 to 83 are automatic discrimination threshold method processing section 36 for the R image 26, G image 27 and B image 28, respectively. The binarization unit for performing binarization processing with the output values of ~ 38, 84 to 86 are R binary images, G binary images and B binary images held in a memory or a work area (not shown), and 87 are Y signal binary image 32, R binary image 84,
A with each of the G binary image 85 and the B binary image 86
AND operation synthesis section for executing ND operation, 88 to 90,
YR composite binary image, YG composite binary image and YB composite binary image stored in a memory or a work area (not shown), 9
1 determines the black pixel frequency in the histogram of each of these composite binary images, and determines whether or not a single dropout color type can be specified based on the black pixel frequency (see step (s12) in FIG. 14). The black pixel frequency calculation / determination unit 92 selects a plane image (red plane image in the case shown) specified by the black pixel frequency calculation / determination unit 91 and outputs the selected plane image. An integrated block composed of components 81 to 91 is shown.

FIGS. 7 to 9 are explanatory views showing an outline (part 3) of the dropout plane detection method. Figure 7 is R
In the case of the plane, FIG. 8 shows the case of the G plane, and FIG. 9 shows the case of the B plane.

In FIGS. 7 to 9, 17 and 51 are
Components having the same contents as the corresponding numbers in FIGS. 3 to 5, 5
2 ′ is a local binarization method (FIGS. 17 to 19) for the red plane of the read form using the automatic discrimination threshold (= 123).
The first red plane binary image obtained by binarizing
53 ′ is a second red plane binary image obtained by binarizing the red plane of the read form by the local binarization method using the blur correction threshold value (= 159), and 56 is the Y signal X of the binary image 51 and the first red plane binary image 52 '
The OR composite image (black pixel frequency = 5.943%), 57 is
Y signal binary image 51 and second red plane binary image 53 '
XOR composite image with (black pixel frequency = 6.014%), 6
Reference numeral 2'denotes a first green plane binary image obtained by binarizing the green plane of the reading form by the local binarization method using the automatic discrimination threshold (= 103), and 63 'indicates the reading form. The second green plane binary image 66 obtained by binarizing the green plane of No. 2 by the local binarization method using the blur correction threshold value (= 133) is the Y signal binary image 51 and the second green plane binary image 66. 1
XOR composite image (black pixel frequency = 4.424%) with the green plane binary image 62 ′ of No.
XOR composite image (black pixel frequency = 3.577%) of the second green plane binary image 63 ′ and the second green plane binary image 63 ′ is the local plane using the automatic discrimination threshold (= 111) for the blue plane of the read form. The first blue plane binary image 73 ′ obtained by binarization by the dynamic binarization method is the local binarization method using the blur plane correction threshold (= 144) of the blue plane of the read form. Second blue plane binary image obtained by binarizing with,
6 is an XOR composite image of the Y signal binary image 51 and the first blue plane binary image 72 '(black pixel frequency = 4.404).
, 77 is an XOR composite image of the Y signal binary image 51 and the second blue plane binary image 73 '(black pixel frequency = 4.%).
219%), respectively. The black level is set to the logical value "1".

The detection method of FIGS. 7 to 9 is the binary image of each plane of the read form and the Y obtained from each plane image.
It focuses on the difference in black pixel frequency between the signal binary image (luminance signal image) and the XOR combined image.

XOR composite images 56, 57, 66, 67,
The black pixel frequencies of 76 and 77 are in a so-called inverse relationship to the respective ones in the AND combined images of FIGS.

That is, in the case of these XOR composite images,
Of the black pixels of the Y signal binary image, the white pixel portions (pixel portions that have disappeared by the binarization process) of the red, blue, and green plane binary images are inverted to black pixels, while The black pixel portion (pixel portion that has not disappeared due to the binarization process) of the value image is inverted to the white pixel.

Therefore, the black pixel frequency of the XOR composite image of each plane becomes maximum in the red plane in which the part that disappears (becomes white) by the binarization processing of the plane image is larger than the other planes.

XOR composite images 56, 57 of each plane,
The black pixel frequencies at 66, 67, 76, and 77 are (1) in the case of the red plane of FIG. 7: 5.943% (automatic determination threshold: 123) 6.014% (blurring correction threshold: 159 = 123 ×
1.3) (2) In the case of the green plane of FIG. 8, -4.424% (automatic determination threshold value: 103) -3.577% (blurring correction threshold value: 133 = 103x)
1.3) (2) In the case of the blue plane in FIG. 9, 4.404% (automatic determination threshold: 111), 4.219% (blurring correction threshold: 144 = 111 ×)
It was 1.3). It should be noted that these numerical values are values in the entire image of each plane (not only the image of the ¼ portion shown in the drawing).

By obtaining and comparing the black pixel frequencies in the above XOR images of the respective red, blue, and green planes, the plane with the lowest frequency (the red plane in the case of the illustrated form) is the dropout plane. Can be specified as

The number of pixels corresponding to the black pixel frequency difference between the XOR composite images of each of the planes is A5 as described above.
In the reading form with the total number of pixels of 1,557,790 pixels in size, (1) In the case of the red plane and the green plane, 23,662 pixels (1.519% of the total: when the automatic discrimination threshold value), 37,963 pixels ( 2.437% of the whole: at the blur correction threshold) (2) In the case of the red plane and the blue plane: 23,974 pixels (1.539% of the whole: at the automatic discrimination threshold), 27,962 pixels (1.795% of the whole: when the threshold value for blur correction) (3) In the case of the green plane and the blue plane: 311 pixels (0.02% of the whole: when the automatic discrimination threshold value); 10,001 pixels ( 0.642% of the whole: when it is a threshold value for blur correction).

FIG. 10 is an explanatory diagram showing the overall structure corresponding to the detection method of FIGS.

In FIG. 10, 21 to 38 are constituent elements having the same contents as the corresponding numbers in FIGS. 2 and 6, and 101.
Nos. 103 to R images 26, G images 27 and B images 28
The local binarization units 104 to 106 that binarize each of the above by the local binarization method (see FIGS. 17 to 19) to create each plane binary image, 104 to 106 are memory and work areas (not shown). The R binary image, the G binary image, and the B binary image 107 held in the table are the Y signal binary image 32 and the R binary image 1
04, an XOR operation synthesizing unit for executing an XOR operation with each of the G binary image 105 and the B binary image 106, 10
8 to 110 are YR composite binary images, YG composite binary images and YB composite binary images held in a memory or a work area (not shown), and 111 is a black pixel frequency in the histogram of each of these composite binary images. And determines whether a single dropout color type can be identified based on the black pixel frequency (see step (s13) in FIG. 14). The black pixel frequency calculation / determination unit 112 calculates the black pixel frequency. A plain image selection / output unit that selects and outputs a plain image (a red plane image in the case shown) identified by the determination unit 111 and an integrated block 113 that includes components 101 to 111, respectively.

The local binarization units 101 to 103 have a function of adjusting the local binarization target range shown in FIG.

FIG. 11 is an explanatory view showing an example of a form having blue and red dropout colors, and 121 is
A blue-based dropout color stamp frame and 122 represent a red-based dropout color entry character frame, respectively.

FIG. 12 is an explanatory diagram showing a method of outputting a composite image for the form shown in FIG.

In FIG. 12, 131 is Y of the read form.
Y generated by binarizing the signal image by the local binarization method
A signal binary image 132 is an R plane binary image generated by binarizing the red plane image of the reading form with an automatic discrimination threshold, and 133 is an AND of the Y signal binary image 131 and the R plane binary image 132. The YR composite binary image, 134 generated by arithmetic composition is a B plane binary image, 1 generated by binarizing the blue plane image of the reading form with an automatic discrimination threshold.
Reference numeral 35 indicates a YRB composite binary image generated by AND operation composition of the YR composite binary image 133 and the B plane binary image 134, respectively. The black level has a logical value of "1".

Here, the red dropout color (entry character frame 122) disappears from the YR composite binary image 133, and the blue series in addition to the red dropout color from the YRB composite binary image 135. Dropout color (seal frame 1
21) also disappears.

The YRB composite binary image 135 and the G plane binary image (not shown) are also combined by AND operation, but in the case of the read form shown, the G color portion is positively included. Since it does not exist, there is no particular change in the image due to this combining process.

FIG. 13 is an explanatory diagram showing the overall structure corresponding to the output method of FIG.

In FIG. 13, 21-32, 39, 4
4, 91, 93, 111 and 113 are components having the same contents as the corresponding numbers in FIG. 2, FIG. 6 and FIG.
Reference numeral 41 denotes an RGB binary image selection unit that sequentially selects the R binary image 104, the G binary image 105, and the B binary image 106 in the integrated block 113 one by one (the order is arbitrary),
142 is an AN of the Y signal binary image 32 and, for example, the R binary image 104 selected by the RGB binary image selection unit 141.
An AND operation synthesis unit that executes D operation, 143 is a YR synthesis binary image held in a memory or a work area (not shown), and 144 is selected by the YR synthesis binary image 143 and the RGB binary image selection unit 141. For example, the G binary image 105
An AND operation combining unit 145 performs an AND operation with the YRG composite binary image 145 held in a memory or a work area (not shown), and a reference numeral 146 represents the YRG composite binary image 145 and RG.
For example, an AND operation combining unit 147 that performs an AND operation with the B binary image 106 selected by the B binary image selecting unit 141.
Is the YRG stored in the memory or work area (not shown).
B composite binary image, 148 is a YRGB composite binary image 14
7 shows a composite binary image output unit that outputs 7 respectively.

FIG. 14 is an explanatory diagram showing a processing procedure corresponding to the overall configuration of FIG. 13, and the contents thereof are as follows. (s1) The target form is read by the two-dimensional color CCD and the process proceeds to the next step. (s2) Divide this read form into RGB planes,
Go to the next step. (s3) A Y signal binary image of the read form is generated based on each of the plane images, and the process proceeds to the next step. (s4) The histogram calculation of each RGB plane image is executed, and the process proceeds to the next step. (s5) The automatic thresholds (Th1 to Th3) in each of the plane histograms are calculated (see FIG. 16), and the process proceeds to the next step. The calculated automatic threshold values themselves and Th1 to Th3 are associated with each other in the form of Th1 ≧ Th2 ≧ Th3. (s6) In each of the automatic thresholds, it is determined whether or not the conditions of Th1-Th2 ≧ 30 and Th2-Th3 ≦ 10 are both satisfied. If “satisfies”, proceed to the next step, and if not, step Proceed to (s9). The determination subject at this time is the plain image determination / selection unit 39. (s7) The plane image (red plane in the case of the form in FIG. 1) corresponding to the automatic threshold Th1 is binarized, and the process proceeds to the next step. (s8) The AND binary combination of this plane binary image and the previous Y signal binary image is executed, and the process proceeds to step (s20). (s9) Each RGB plane image is binarized, and the process proceeds to the next step. (s10) This AND binary combination of each plane binary image and the previous Y signal binary image is executed, and the process proceeds to the next step. (s11) Black pixel frequency (Bf1 to Bf) in each composite image
f3) is calculated, and the process proceeds to the next step. The calculated black pixel frequency itself and Bf1 to Bf3 are
It corresponds in the form of Bf1 ≦ Bf2 ≦ Bf3. The black pixel frequency is calculated as the ratio of the number of black pixels to the total number of pixels of the composite image. (s12) At each black pixel frequency, it is judged whether or not the conditions of Bf2-Bf1 ≧ 1.5% and Bf3-Bf2 ≦ 0.2% are both satisfied. If “satisfies”, the process proceeds to step (s19). , If not satisfied, go to the next step. The determination subject at this time is the black pixel frequency calculation / determination unit 91. (s13) X of the previous plane binary image and Y signal binary image
The OR operation composition is executed, and the process proceeds to the next step. (s14) Black pixel frequency (Bf4 to Bf) in each composite image
f6) is calculated and the process proceeds to the next step. The calculated black pixel frequency itself and Bf4 to Bf6 are
Correspond in the form of Bf4 ≧ Bf5 ≧ Bf6. (s15) At each black pixel frequency, it is judged whether or not the conditions of Bf4-Bf5 ≧ 1.0% and Bf5-Bf6 ≦ 0.1% are both satisfied, and if “satisfies”, proceed to step (s19). , If not satisfied, go to the next step. The determination subject at this time is the black pixel frequency calculation / determination unit 111. (s16) A of the Y signal binary image and the R plane binary image
ND arithmetic synthesis is executed and the process proceeds to the next step. (s17) This YR composite binary image and the preceding G plane binary image are ANDed and composited to proceed to the next step. (s18) This YRG composite binary image and the previous B plane binary image are ANDed and composited, and the process proceeds to step (s20). (s19) An image corresponding to the black pixel frequency Bf1 or Bf4 is selected, and the process proceeds to the next step. The corresponding image of Bf1 is a synthetic binary image (YR synthetic binary image in the case of the form of FIG. 3), and the corresponding image of Bf4 is a plain binary image (a red plane binary image in the case of the form of FIG. 7). ). (s20) The composite image of step (s18) and the selected image of step (s19) are output to a display screen (not shown) or the like.

In FIG. 13 and FIG. 14, all the block processes of the integrated blocks 44, 93 and 113 may not be executed, and only one or two processes of the integrated block may be performed. In addition, integrated block 4
The execution order of the processes of 4, 93, and 113 is also arbitrary. For example, the order of the integrated blocks 93-113-44 may be used.

It is needless to say that the constituent elements of the integrated block change according to such a change in the processing mode.
For example, when the integrated block 44 is omitted, the histogram calculation units 33 to 33 are included as constituent elements of the integrated block 93.
35 and automatic discrimination threshold method processing units 36 to 38 are added.

In the above form image processing, the Y signal binary image is mainly generated by the local binarization method, and the red, blue, and green plane binary images are automatically discriminated by the threshold value method and the local binary method. It is generated by the binarization method.

FIG. 15 is an explanatory view showing the outline of the CCD and pixel shift system. FIG. 15A shows the arrangement of the photoelectric conversion elements of the CCD, and FIG. 15B shows the shift of the image pickup pixels.

As shown in the figure, the photoelectric conversion portion for one pixel of the CCD comprises one red conversion element R, one blue conversion element B and two green conversion elements G.

When a form is read by a CCD in which 400,000 photoelectric conversion units (4 elements of RGGB) are arranged at a predetermined interval (two circles), the pixel signal of the (1) part is first detected by the first image pickup. I want it.

Thereafter, the CCD or the form is shifted eight times in the horizontal direction and the downward direction in the drawing to sequentially obtain the pixel signals of the respective portions (2) to (9).

By combining the individual pixel signals individually obtained by these nine imagings, 3.6 million pixels (= 400,000 ×)
The image signal of 9) is obtained.

FIG. 16 shows a multivalued plane image (0 to 255).
It is explanatory drawing which shows the process sequence which calculates | requires automatic discrimination thresholds, such as a gradation, and the content is as follows. (s31) The normalized histogram value pi at the density level i (0 ≦ i ≦ 255) of each gradation is calculated by “pi = ni / N”. Note that ni is the number of pixels of the density level i, and N is the number of read pixels of the entire form. (s32) The average density level μt of this normalized histogram is calculated by the equation shown. (s33) p from gradation 0 to gradation k (0 ≦ k ≦ 255)
Cumulative value ω (k) of i and cumulative value m of “i * pi”
(K) is calculated for each gradation according to the illustrated formula. (s34) Using the cumulative values ω (k), m (k) and the average density level μt, the variance value σ (k) corresponding to each gradation k (0 ≦ k ≦ 255) is calculated by the formula shown. (s35) The gradation k at which the variance value σ (k) is maximum is set as the binarization threshold value of the multi-valued plane image.

FIG. 17 is an explanatory view showing the outline of the local binarization method.

In FIG. 17, reference numeral 151 is a target pixel of the local binarization processing, 152 is a 7 × 7 pixel rectangular filter for the target pixel, 153 is a character portion stroke pixel in the rectangular filter, and 154 is a rectangle. Background pixel in the filter, σmin is a threshold value for background separation in the rectangular filter (for example, 100), E is a density average value of each pixel in the rectangular filter, and E2 is a density mean square value of each pixel in the rectangular filter. , G2 are obtained by the equation shown using the density average value E and the density mean square E2, and S is the density dispersion value in the rectangular filter, and S is obtained by the equation shown using the density average value E and the density dispersion value g2. The binarization threshold value of the target pixel 151, which is obtained, is shown.

The threshold value S for binarization is used when the rectangular filter 152 is judged not to be in the background portion.
That is, when g2> σmin holds.

In the local binarization method, a rectangular filter 152 including the target pixel for binarization is set,
By obtaining the density dispersion value g2 of all the pixels in it and examining the magnitude relationship between it and the background separation threshold g2,
It is determined whether the rectangular filter corresponds to the background portion or the information portion.

When it is determined to be the background portion, "white" and "black" of the target pixel 151 are specified based on the magnitude of the density dispersion value g2 (without directly using the density of the target pixel).

On the other hand, when it is determined that the information part is “white”, it is determined based on the density level of the target pixel 151.
Identify "black".

FIG. 18 is an explanatory diagram showing the processing procedure of the local binarization method, and the contents thereof are as follows. (s41) The density average value E of the rectangular filter 152 is calculated. (s42) The density mean square value E2 of the rectangular filter 152 is calculated. (s43) Density dispersion value g2 of the rectangular filter 152 (= E2-
E * E) is calculated. (s44) When “the threshold for background separation σmin <density dispersion value g2”, the rectangular filter 152 is determined to be the background portion and the process proceeds to the next step. When the magnitude relation is not established, the rectangular filter 152 is set to the non-background. Step (s4
Proceed to 7). Σmin is an initial setting value (value is arbitrary)
And is set to “100” here. (s45) The threshold value for binarization S (= E + (g2) ^1/2 * 0.1) is calculated. By reducing the multiplication coefficient, it is possible to suppress noise in the pixels around the character. (s46) If "binarization threshold value S <density of target pixel 151", proceed to step (s48). If the magnitude relation is not satisfied, proceed to step (s49). (s47) If "reference value <average density E", proceed to the next step, and if the magnitude relationship is not satisfied, proceed to step
Proceed to (s49). Note that what is used as the reference value is arbitrary, and for example, the automatic determination threshold value of the target image may be used as the reference value. (s48) The target pixel 151 is determined to be "white". (s49) The target pixel 151 is determined to be “black”.

FIG. 19 is an explanatory diagram showing adjustment of a local binarization target range (a range estimated to be a character blurred portion), where ot is an automatic discrimination threshold in the target image histogram, and a and b are local. The coefficient for adjusting the target binarization range, b-ave is the density average of the black area (density 0 to ot), w-ave is the density average of the white area (density ot + 1 to 255), and bt is Lower limit concentration of local binarization target range, wt
Indicates the upper limit concentration of the local binarization target range, respectively.

Here, the local binarization target range (density bt to wt) in the target image histogram is calculated by the two equations shown. The local binarization target range can be arbitrarily set by adjusting the coefficients a and b in the equation. The adjustment of the coefficients a and b is performed by the operation of hardware or the execution of software in the local binarization unit.

By executing the local binarization method only on pixels in such an estimated range (density bt to wt), a significantly high speed processing is possible.

Each pixel outside the estimation range is uniformly determined as "black (density 0 to bt)" or "white (density wt to 255)". With respect to the blurring of letters and numbers in this white area, it is possible to determine that this pixel portion is “black” by increasing wt. On the other hand, as the estimation range is expanded, the time required for binarization of the entire image becomes longer.

In this example, "a = 0.35, b = 0.6
When the local binarization method was executed for 5 to 8% of the pixels of the entire image by setting to "2", a binarized image of the same level as that in the case of locally binarizing all the pixels was obtained.

FIG. 20 is an explanatory diagram showing an example of change in a binary image due to blur correction during local binarization. FIG. 20A shows the background separation threshold value in FIG. 17 changed (σmin = 100 → 5).
0), and FIG. 19B shows the local binarization target range of FIG. 19 changed (a = 0.35, b = 0.62 → a = 0.
35, b = 0.90).

Recording media for storing the processing programs for detecting the dropout plane and outputting the composite image are as follows: database on the program provider side (destination memory such as DASD), portable recording media of various formats It may be a RAM or a hard disk on the computer body side. The program is loaded into the computer main body and executed on its main memory.

(Supplementary Note 1) Density for obtaining a pixel density distribution for each plane of the read image in a form reading device that optically reads a form including a dropout color portion and a user entry portion related thereto in pixel units. A form reading apparatus comprising: a distribution calculating means; and a plane selecting means that obtains a characteristic value of the pixel density distribution and specifies a dropout plane based on the characteristic value. (Supplementary Note 2) The form reading device according to Supplementary Note 1, wherein the density distribution calculating means uses an automatic determination threshold value of the pixel density distribution as the characteristic value. (Supplementary Note 3) In a form reading device for optically reading a form including a dropout color part and a user entry part related thereto, in a form reading device, a binarizing means for binarizing each plane of the read image. A form reading device comprising: a plane selecting unit that obtains a black pixel frequency or a white pixel frequency corresponding to the plane binary image and specifies a dropout plane based on the pixel frequency. (Supplementary Note 4) The form reading device according to Supplementary Note 3, wherein the binarizing means has a binarization processing correction function for dealing with a blurred portion of a character / number written by a user. (Supplementary Note 5) The plane selection means uses, as the black pixel frequency or the white pixel frequency, a pixel frequency of a composite binary image generated by a calculation process of a luminance signal binary image of the read image and the plane binary image. The form reading device according to appendix 3, characterized in that. (Supplementary Note 6) In a form reading device that optically reads a form including a dropout color part and a user entry part related thereto in pixel units, a plane binary image and a luminance signal binary image of the read image are obtained. After generating the first combined binary image by the image generation means for generating and the arithmetic processing of the luminance signal binary image and one of the plane binary images, the first combined binary image and the A second combined binary image is generated by an arithmetic operation with the remaining one of the plane binary images, and an arithmetic processing with the second combined binary image and the other one of the plane binary images is performed. And a image combining means for generating a third combined binary image according to the above. (Supplementary Note 7) The form reading apparatus according to Supplementary Note 6, wherein the image generating means has a binarization correction function for dealing with a blurred portion of a character / number written by a user. (Supplementary Note 8) In a program used in a process of reading a form including a dropout color part and a user entry part related thereto, the program obtains a characteristic value of pixel density distribution of each plane of the form read image, A form reading processing program for causing a computer to realize a function of specifying a dropout plane based on a characteristic value. (Supplementary Note 9) In a program used in a process of reading a form including a dropout color part and a user-written part related thereto, the program is a black pixel frequency corresponding to each plane binarized image of the form read image, or A form reading program for causing a computer to realize a function of obtaining a white pixel frequency and specifying a dropout plane based on the pixel frequency. (Supplementary Note 10) In a program used in a process of reading a form including a dropout color portion and a user entry portion related thereto, the program is a luminance signal binary image and the plane binary image of the form reading image. To generate a first combined binary image, and then calculate an operation of the first combined binary image and the remaining one of the plane binary images into a second combined binary image. A computer for realizing a function of generating an image and of generating a third combined binary image by the arithmetic processing of the second combined binary image and the remaining one of the plane binary images A form reading processing program characterized by:

[0093]

As described above, according to the present invention, the characteristic value (for example, the automatic discrimination threshold value) of the pixel density distribution in each plane of the read image of the target form, the black pixel frequency or the white value corresponding to each plane binary image, or Based on the pixel frequency, that is, without performing combining processing such as AND operation between plane binary images,
Since the dropout plane is specified, it is possible to efficiently execute the generation process of the form image in the form of removing the dropout color.

Further, since the black pixel frequency or the white pixel frequency of the composite binary image (having higher resolution than the plane binary image itself) of the luminance signal binary image of the read image and the plane binary image is used, The accuracy of identifying the dropout plane can be improved.

Further, when the dropout plane cannot be specified based on the above-mentioned characteristic value or black pixel frequency, when the composite binary image of each plane is generated, arithmetic composition with the luminance signal binary image of the form is also executed. Therefore, the resolution of the composite binary image itself can be increased.

In addition, the threshold for binarization, the threshold for background separation of the local binarization method, the target range, etc. are adjusted so that the faint portions of the characters / numbers entered by the user are appropriately converted into black pixels. Therefore, it is possible to improve the accuracy of identifying the dropout plane and to increase the recognition accuracy of the characters and numbers in the final output image by making them thicker.

[Brief description of drawings]

FIG. 1 is an explanatory diagram showing an outline (part 1) of a dropout plane detection method.

FIG. 2 is an explanatory diagram showing the overall configuration corresponding to the detection method of FIG.

FIG. 3 is an explanatory diagram showing an outline (part 2: R plane) of a dropout plane detection method.

FIG. 4 is an explanatory diagram showing an outline (part 2: G plane) of a dropout plane detection method.

FIG. 5 is an explanatory diagram showing an outline (part 2: B plane) of a dropout plane detection method.

FIG. 6 is an explanatory diagram showing an overall configuration corresponding to the detection method of FIGS.

FIG. 7 is an explanatory diagram showing an outline (part 3: R plane) of a dropout plane detection method.

FIG. 8 is an explanatory diagram showing an outline (part 3: G plane) of a dropout plane detection method.

FIG. 9 is an explanatory diagram showing an outline (part 3: B plane) of a dropout plane detection method.

FIG. 10 is an explanatory diagram showing an overall configuration corresponding to the detection method of FIGS.

FIG. 11 is an explanatory diagram showing an example of a form having blue-based and red-based dropout colors.

12 is an explanatory diagram showing an outline of a method of outputting a composite image for the form of FIG.

FIG. 13 is an explanatory diagram showing an overall configuration corresponding to the output method of FIG.

FIG. 14 is an explanatory diagram showing a processing procedure corresponding to the overall configuration of FIG.

FIG. 15 is an explanatory diagram showing an outline of a CCD and pixel shifting method.

FIG. 16 is an explanatory diagram showing a processing procedure of an automatic discrimination threshold method of a plane multi-valued image.

FIG. 17 is an explanatory diagram showing an outline of a local binarization method (filter).

FIG. 18 is an explanatory diagram showing a processing procedure of a local binarization method.

FIG. 19 is an explanatory diagram showing adjustment of a local binarization target range.

FIG. 20 is an explanatory diagram showing an example of changes in a binary image due to blur correction during local binarization.

[Explanation of symbols]

11: Red plane 12: Red plane multi-valued image histogram 13: Green plane 14: Histogram of green plane multi-valued image 15: Blue plane 16: Histogram of blue plane multi-valued image 17: Red frame on the form (dropout part) 21: Two-dimensional color CCD 22: Pixel shifting means 23: Captured image synthesizing means 24: Shading correction means 25: R, G, B plane dividing means 26-28 : Red, green, and blue plain images 29: Y signal image generating means, 30: Y signal image 31: Local binarization unit 32: Y signal binary image 33-35 : Histogram calculator 36-38 : Automatic discrimination threshold method processing unit 39: Plain image determination / selection unit 40: Selected image binarization unit 41: Binary image of selected plane 42: AND operation synthesis section 43: Composite image output unit 44: Integrated block 51: Y signal binary image 52,52 ' : First red plane binary image 53,53 ' : Second red plane binary image 54: AND composite image (black pixel frequency = 11.823%) 55: AND composite image (black pixel frequency = 12.030%) 56: XOR composite image (black pixel frequency = 5.943%) 57: XOR composite image (black pixel frequency = 6.014%) 62, 62 ': first green plane binary image 63, 63 ': second green plane binary image 64: AND composite image (black pixel frequency = 13.653%) 65: AND composite image (black pixel frequency = 15.495%) 66: XOR composite image (black pixel frequency = 4.424%) 67: XOR composite image (black pixel frequency = 3.577%) 72, 72 ': first blue plane binary image 73, 73 ': second blue plane binary image 74: AND composite image (black pixel frequency = 13.787%) 75: AND composite image (black pixel frequency = 16.273%) 76: XOR composite image (black pixel frequency = 4.404%) 77: XOR composite image (black pixel frequency = 4.219%) 81-83: Binarization unit 84: R binary image 85: G binary image 86: B binary image 87: AND operation composition section 88: YR composite binary image 89: YG composite binary image 90: YB composite binary image 91: Black pixel frequency calculation / determination unit 92: Plain image selection / output unit 93: Integrated block 101-103: Local Binarization Unit 104: R binary image 105: G binary image 106: B binary image 107: XOR operation synthesis section 108: YR composite binary image 109: YG composite binary image 110: YB composite binary image 111: Black pixel frequency calculation / determination unit 112: Plain image selection / output unit 113: Integrated block 121: Seal frame (blue dropout color) 122: Entry character frame (red dropout color) 131: Y signal binary image 132: R plane binary image 133: YR composite binary image 134: B plane binary image 135: YRB composite binary image 141: RGB binary image selection unit 142: AND operation synthesis section 143: YR composite binary image 144: AND operation synthesis section 145: YRG composite binary image 146: AND operation synthesis section 147: YRGB composite binary image 148: Synthetic binary image output unit 151: Target Pixel for Local Binarization 152: 7 × 7 pixel rectangular filter for the target pixel 153: Stroke pixel of character part 154: Background pixel σmin: threshold value for background separation (for example, 100) E: Density average value of each pixel in the rectangular filter E2: density mean square value of each pixel in the rectangular filter g2: Density dispersion value in the rectangular filter S: Threshold for binarization of target pixel ot: automatic discrimination threshold a, b: Coefficients for adjusting the local binarization target range b-ave: average density of black area (density 0 to ot) w-ave: average density of white areas (density ot + 1 to 255) bt: lower limit concentration of local binarization target range wt: upper limit concentration of local binarization target range

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Yasunao Izaki             4-1, Kamiodanaka, Nakahara-ku, Kawasaki-shi, Kanagawa             No. 1 within Fujitsu Limited (72) Inventor Shunji Sakane             4-1, Kamiodanaka, Nakahara-ku, Kawasaki-shi, Kanagawa             No. 1 within Fujitsu Limited (72) Inventor Katsumi Ide             4-1, Kamiodanaka, Nakahara-ku, Kawasaki-shi, Kanagawa             No. 1 within Fujitsu Limited (72) Inventor Kazumasa Komatsu             1-15-3 Dogenzaka, Shibuya-ku, Tokyo Rye             Fu Consultant Co., Ltd. F-term (reference) 5B029 BB02 DD04

Claims (5)

[Claims] 1. A density distribution calculation for obtaining a pixel density distribution for each plane of a read image in a form reading device that optically reads a form including a dropout color part and a user entry part related thereto in pixel units. A form reading apparatus comprising: a means for determining a characteristic value of the pixel density distribution; and a plane selecting means for specifying a dropout plane based on the characteristic value. 2. A form reading device for optically reading a form including a dropout color portion and a user entry portion related thereto in a pixel unit, and binarizing means for binarizing each plane of the read image. And a plane selecting means for determining a black pixel frequency or a white pixel frequency corresponding to the plane binary image and specifying a dropout plane based on the pixel frequency. 3. A form reading device for optically reading a form including a dropout color part and a user-written part related thereto in pixel units, wherein each plane binary image and luminance signal binary image of the read image And an image generating means for generating a first combined binary image by calculating the brightness signal binary image and one of the plane binary images. A second combined binary image is generated by an operation process with the remaining one of the plane binary images, and an operation between the second combined binary image and the remaining one of the plane binary images is performed. A form reading apparatus comprising: an image synthesizing unit that generates a third synthetic binary image by processing. 4. A program used in a process of reading a form including a dropout color part and a user entry part related to the dropout color part, wherein the program obtains a characteristic value of pixel density distribution of each plane of the form read image, It is for realizing the function of identifying the dropout plane based on the characteristic value,
A form reading processing program characterized by the above. 5. A program used in a process of reading a form including a dropout color part and a user entry part related to the dropout color part, wherein the program is a black pixel frequency or black pixel frequency corresponding to each plane binary image of the form read image. A form reading program for causing a computer to realize a function of obtaining a white pixel frequency and specifying a dropout plane based on the pixel frequency.