JPH06113129A - Color picture processor - Google Patents

Abstract

PURPOSE:To efficiently update reference data for detecting a specified original. CONSTITUTION:A color picture processor having a means for detecting the presence or absence of the specified original by an input color signal is provided with the registration means (color space matching circuit 109) of the specified original, a picture read means (CCD 101 and others) reading the specified original by plural different conditions and a means (CPU 112) registering the specified original from plural pieces of picture data which are read.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a color image processing apparatus for detecting a specific object in a document image from an input color signal.

[0002]

2. Description of the Related Art In recent years, with the development of color copying machines, the reproducibility of original copying has been improved. Therefore, a technique for preventing counterfeiting of bills, securities, etc. is required.
As one of such techniques, data of a specific original document in the RGB color space is registered in advance, and it is determined whether or not the color distribution of the input image data matches the distribution of the specific original document data in the color space to specify The applicant has proposed a technique for discriminating a manuscript.

More specifically, R (red) and G (green) of input original image data, R and G of B (blue) signals,
B. The distribution in the three-dimensional coordinate space is compared with the distribution of the specific document data in the three-dimensional coordinate space, and if the matching part is a certain value or more, the specific document is included in the documents to be copied. It was to determine that it was.

In the above arrangement, the specific original to be judged must have the distribution of the tint in the R, G and B color spaces registered in advance. However, there are various types of specific originals that are not suitable for copying, and they greatly differ depending on the country, region, occupation, etc. in which the color copying machine is used.
Therefore, it is impossible to register all the data of those manuscripts, and it is desirable that the registration change or new registration work can be easily performed depending on the usage environment.

Therefore, in order to efficiently execute such a registration operation of the specific original by the applicant, the determination parameter of the determination parameter is determined based on the image data obtained by reading the specific original as a target with a color copier. A method of setting is proposed.

[0006]

However, in the above-mentioned conventional example, since there is only one specific original to be read, the accuracy of the determination is likely to change depending on, for example, the dirt of the specific original or the place on the original table. was there.

[0007]

In order to solve the above-mentioned problems, therefore, in the present invention, an original image reading means for reading a specific original to be judged under a plurality of different conditions and a plurality of read image data are provided. By providing a storage unit for storing, a unit for generating a determination parameter for identifying a specific document from the stored image, and a unit for storing the parameter, the presence or absence of an arbitrary specific document is determined as a stain on the specific document. It is possible to make a judgment with high accuracy without being affected by the above.

[0008]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a signal processing block diagram of a color image processing apparatus. In the figure, 101 is a CCD color sensor which reads a color original (not shown) line by line and outputs read data to an analog amplifier 102. 1
Reference numeral 03 is an A / D converter, which samples and holds the analog output of 102 and converts it into digital data. Reference numeral 104 denotes a shading correction circuit that corrects the inclination of brightness depending on the reading position of the image signal and the sensitivity unevenness of the CCD.

Reference numerals 105 to 107 are a logarithmic conversion circuit for converting an RGB signal into a YMCK signal for printer output by known color correction processing, a black component extraction circuit, and a masking / UCR (undercolor removal) circuit. Reference numeral 108 denotes a circuit that generates read synchronization signals HS, CLK, and VS, and this synchronization signal is supplied to all circuit blocks. Here, HS is a main scanning section signal, CLK is a pixel reading basic clock signal, and VS is a section signal indicating a sub-scanning direction effective area. Reference numeral 109 is a color space matching circuit that calculates the similarity of the read image data and the specific original in the color space in real time. Reference numeral 110 is 1 when it is determined that the specific original exists based on the matching result of 109, and 0 is output otherwise. Is a determination signal generating circuit. The judgment signal output of 110 is input to, for example, the masking / UCR circuit of 107, and if a specific document exists, the masking coefficient is changed from a normal value to deteriorate the color reproducibility of the copy output, resulting in a forgery action. Can be brought to an undesired state.

Reference numeral 111 denotes a frame memory for storing a plurality of screens of an input original image, which can store the input image and send the stored image data to a color space matching circuit 109 to perform a determination operation. 112 is 111
This is a CPU for accessing the data in the frame memory, and also writes the data in 109 and 110. 1
Details of 11, 112 will be described later.

(Color Space Matching Circuit) FIG. 2 is a block diagram for explaining the color space matching circuit 109. In the figure, R201 is the data of the upper 5 bits of the 8 bits of the R (red) signal from the shading correction circuit 104. Similarly, G202 and B203 are high-order 5-bit data of G (green) signal and B (blue) signal.

Reference numeral 204 is an EE-PROM (readable / writable / erasable memory) in which information regarding the tints of a plurality of types of specific originals is stored. The R, G, and B signals are input to the addresses A _{0 to} A ₁₄ , and a determination signal indicating whether or not the R, G, and B signals match the tint of the specific document is included in the data D _{0 to} D ₇ . Is output. The ROM 204 stores information about the tint of a specific original, and is 1 when the input R, G, B signals match the tint of the specific original.
However, if not, 0 is output to D _{0 to} D ₇ .

FIG. 4 is a diagram showing the relationship between the data relating to a plurality of originals stored in the ROM 204 and the bit positions in the ROM 204. As a result, from D _{0 to} D ₇ ,
Judgment information (0 or 1) regarding the tint of eight different types of specific originals (specific original A to specific original H) is output in parallel.

Reference numerals 210 to 217 denote tint determination signals X ₀ to
This is a circuit for performing the smoothing operation shown in FIGS. 5 and 6 on the signal of X ₇ . FIG. 5 is a block diagram showing a circuit configuration of the smoothing circuits 210 to 217.

FIG. 6 shows the input Xn and the output value Yn after the smoothing process.
It is a figure which shows the relationship with. The value of Yn increases when 1 continues as the value of input Xn. When Xn becomes 0, the Yn value gradually decreases. As a result, when the input R, G, B signals continuously match the tint of the specific original, Xn continuously takes the value of 1, and Yn increases until it exceeds a predetermined threshold value. Output C ₀ -C ₇ of the smoothing circuit is obtained by binarizing the Yn at predetermined threshold, the C ₀ -C ₇ only if the pixel meets the color of the thus specific original is continuous 1 is output, which enables more accurate determination. The color space determination circuits 220 to 227 are RGB
This is a circuit that calculates the similarity between the specific document data in the color space and the input color signal in real time, and outputs the similarity determination signals MK _{0 to} MK ₇ .

FIG. 3 is a block diagram of one of the color space determination circuits 220 to 227. The data Dn from the SRAM 230 and the signal Cn from the smoothing circuit are 303 at O.
The R operation is performed and the result is written again in the SRAM 230 as Dn. Meanwhile data X from Dn, Cn, and 204
n is ANDed at 302, and 1 is output from 302 only when Xn and Cn are 1 and Dn transitions from 0 to 1. When the output of 302 is 1, the counter of 304 is incremented. The counter 304 is cleared at the rising edge of the sub-scanning section signal VS.

Reference numeral 305 is a circuit for holding the maximum output value of the counter 304. The maximum value of the counter output, which is the output of 305, is compared at 307 with a predetermined constant δn. If the maximum value is larger than δn, 307 outputs 1 as MKn, and otherwise outputs 0 as MKn. The value of δn is set as the value of S% of the volume of the tint of the specific document in the color space.

That is, δn = Vorg × S / 100, where Vorg is the number of 1's in the ROM 204 shown in FIG. The value of S is set so that the determination signal MKn becomes 1 when the specific document is placed on the document table and is read up to about half. When the input image data and the data of the specific document have substantially the same shape in the RGB three-dimensional color space by the above processing, the color space similarity determination signals MK _{0 to} MK ₇
Will be 1.

(Determination Signal Generation Circuit) FIG. 7 is a circuit block diagram for explaining the determination signal generation circuit 110. When it is determined that any one of the plurality of specific document data registered in the ROM 204 by the present circuit configuration matches the input image data in the color space, the determination signal f becomes 1 in real time, and the masking / immediation immediately occurs. Feedback to the UCR circuit is performed.

(Judgment Parameter Generation Circuit) The tint distribution data (EEPROM 204 in FIG. 4) and the judgment threshold value (δn in 306 in FIG. 3 and 506 in FIG. 5) required in the above-mentioned specific original judgment circuit. α, etc.) is CPU1
12 determines with reference to the image data of the frame memory 111.

An example of the procedure will be described below with reference to FIG.

FIG. 8 is a flow chart showing the procedure for registering a specific document.

In step S801, one of the eight types of specific document determination circuits shown in FIG. Select the selected number from N (0 to
7).

In S802, the EE-PROM shown in FIG.
Of the contents of 204, the tint information corresponding to the Nth determination circuit is erased.

In step S803, the specific original to be registered is placed on the original table of the image reading apparatus, and is scanned and read.
It is stored in the frame memory 111 as an RGB image.
At this time, if all the original images are stored, the capacity of the frame memory becomes very large. Therefore, for example, input data may be thinned out every 16 pixels and stored.

In this embodiment, the specific original is read under a plurality of different conditions, and all the read image data are once stored in the frame memory. The following may be considered as different conditions. (1) Two specific originals, one with little new stain and the other with old stain, are prepared and read and stored. (2) When a specific document is placed on the platen, two types of images are stored: one is when a white paper is placed behind it to be read, and the other is when a black paper is placed to read it. (3) The place and angle when the specific document is placed on the platen are changed in several ways and read and stored.

In any of the above cases, the same data will not be obtained due to the influence of the color variation of the specific original or the variation of the reading device.

Therefore, if such a data is used to set the judgment parameter by the method as described below, such a variation will be taken into consideration when the judgment of the specific original is actually made. The accuracy will be improved.

In step S804, the existing position of the specific original is detected for one image out of the plurality of images stored in 111. This is because it is not decided where the document (for example, securities, etc.) is placed on the document table, and it is not known which part in the frame memory is meaningful data. This is shown in FIG. In the figure, reference numeral 901 represents the data for one sheet in the frame memory.
02 indicates the portion of the specific document image required.
Therefore, in the data of 901, when the background portion is dark, for example, a pixel whose signal level is equal to or higher than a certain threshold value is detected, and the minimum value (903) and the maximum value (904) of the vertical coordinate and the horizontal direction are detected. If the minimum value (905) and the maximum value (906) of the directional coordinates are obtained, the existence position of the specific document can be limited.

In step S805, the R, G, and B data for one pixel in the specific document existing area obtained above is read from the 111.

In S806, R read in S805,
The color information of the address corresponding to the G and B values is EE-PRO.
Write to M204 (1 is written to the address corresponding to the R, G, and B values in the Nth column shown in FIG. 4).

In S807, the area 90 shown in FIG.
The processing of S805 and S806 is repeated for all pixels of the data of 2. The above procedure from S804 to S807 is repeated for all the plurality of images in the frame memory.

When the processing up to this point is completed, the distribution of the tint of the specific original to be registered, including variations in a plurality of images, is EE.
-It has been registered in the PROM.

In step S808, the maximum value (25 in this case) of the settable thresholds α and δn of the Nth determination circuit is set.
Set to 5).

In step S809, a plurality of document images read in step S803 and stored in the memory are read as N
The second judgment circuit is operated. In FIG. 1, the image data in the frame memory 111 may be sequentially transmitted to the bus connected to the color space matching circuit 109, which is repeated a plurality of times. At this time, if the determination threshold value set in S808 is appropriate, it is detected that registered documents exist for all document images, and 1 should be output as the Nth determination output MKn.

In step S810, if MKn = 1 is output for all the document images in step S809, it is determined that the set threshold value is appropriate, and the registration operation ends (step S811).

If there is even one with MKn = 0, the threshold values are not appropriate, so the values of α and δn are reduced by a predetermined amount, and determination processing S80 for the same document is performed again.
9 is performed (S812).

By repeating the above processing until MKn = 1, all the determination parameters of the target specific original are set.

(Other Embodiments) Although an example of a color copying machine has been described in the above embodiments, an original image input by a scanner or the like is taken into an external computer and the same processing is performed on the computer. It is also possible to write the parameters to be written in the EEPROM and incorporate this ROM into the image processing apparatus of the present invention.

Further, as the judgment parameter, the tint distribution information and the judgment threshold value are used in the above embodiment, but it is of course possible to apply any one of them or another judgment parameter.

As a method of setting the judgment parameter,
Although it may be determined by gradually changing the judgment threshold value as in the first embodiment, it is possible to efficiently use the non-linear optimization method such as the damping least squares method or the back propagation learning rule in the neural network to set the parameter. It is also possible to ask

Further, as another application, the originals to be read are not limited to the specific originals to be judged, and general originals that should not be judged can be read at the same time. That is, when a specific original is determined, when a document image that is very similar to the specific original but is different from the specific original is read, it is mistakenly determined to be the specific original, and the copy output is deteriorated. Inconvenience may occur. Therefore, when the determination parameter is determined, the specific document image and the general document that may be erroneously determined as described above are read so that the determination output MKn = 1 for the specific document, and The determination threshold value is determined so that the determination output MKn = 0 for a general document, and only the specific document is effectively determined, and the general document other than the specific document is erroneously determined. Will be able to prevent.

[0044]

As described above, according to the present invention, it is possible to easily set the parameters of the specific original determination circuit for an arbitrary specific original, so that it is possible to identify a wide variety of specific originals.

Further, variations in the determination accuracy due to variations in element characteristics of the color image processing apparatus and stains on the original image itself are also stable by applying various factors to the determination parameters by applying the present invention. The determination process can be performed.

[Brief description of drawings]

FIG. 1 is an overall block diagram of a color image processing apparatus of the present invention.

FIG. 2 is a diagram illustrating a color space matching determination circuit 109.

FIG. 3 is a diagram illustrating color space determination circuits 220 to 227.

FIG. 4 is a diagram for explaining the contents of EE-PROM 204.

FIG. 5 is a circuit block diagram of smoothing circuits 210 to 207.

FIG. 6 is a diagram illustrating the operation of a smoothing circuit.

FIG. 7 is a circuit block diagram of a determination signal generation circuit 110.

FIG. 8 is a flowchart illustrating a processing procedure of the present invention.

FIG. 9 is a diagram illustrating position detection of a specific document.

[Explanation of symbols]

 109 Color Space Matching Circuit 110 Judgment Signal Generation Circuit 111 Frame Memory 112 CPU

Claims (5)

[Claims] 1. A color image processing apparatus having means for detecting the presence or absence of a specific original from an input color signal, wherein the specific original registration means and the image reading means for reading the specific original under a plurality of different conditions are read. And a means for registering the specific document from a plurality of image data. 2. The registration unit reads image data of a specific original document into the color image processing apparatus under a plurality of different conditions, a storage unit stores a plurality of read images, and a plurality of stored reading units. 2. The color image processing apparatus according to claim 1, comprising a unit for generating a parameter for detecting a specific original from an image, and a unit for storing the parameter. 3. The color according to claim 1, wherein the specific document detecting means comprises means for calculating a degree of similarity between a color distribution of a specific document and a color distribution of an input color signal. Image processing device. 4. The color image processing apparatus according to claim 2, wherein the parameter for detecting the specific document is data characterizing the distribution of the tint of the specific document. 5. The color image processing apparatus according to claim 2, wherein the parameter for detecting the specific document is a determination threshold value for determining the presence or absence of the specific document.