JPH0810901B2 - Color identification device - Google Patents

Description

The present invention relates to a color identification device for identifying a color on a document based on an image signal separated into three colors by a color document reading device.

"Prior Art" In a conventional apparatus for reading a color original, image information on the original is separated into three colors by using a filter or the like,
The three types of image signals thus obtained are binarized. In this device, based on the image signals thus obtained, yellow (Y), cyan (C), magenta (M), red (R), green (G), blue (B)
We distinguish 7 colors, black and black.

"Problem to be Solved by the Invention" In this conventional apparatus, when a threshold for binarization is set in order to accurately identify a certain color, the identification of other colors becomes inaccurate. There was a problem. It was also difficult to increase the number of colors to be identified.

In order to eliminate such a defect, there has been proposed an apparatus which performs three-dimensional mapping by assigning the signal levels of the image signals separated into three colors to the X, Y and Z3 axes. However, in this apparatus, since the colors are assigned three-dimensionally, there is a problem in that the boundary designation of each color becomes complicated. Further, this latter device has a problem that the configuration of the device becomes expensive because many signal lines for performing mapping are required.

Therefore, an object of the present invention is to provide a color identification device capable of accurately identifying colors with a relatively simple configuration.

"Means for Solving Problems" The color identification device of the present invention separates image information on a document into three colors and reads them, and based on the three types of image signals obtained thereby, the color of the image information is obtained. In the color identification device for identifying
A first subtractor for subtracting the signal levels of any two types of image signals of the two types of image signals, and two types of image signals of the two types of image signals used in the first subtractor. A second subtractor for subtracting the signal level of the image signal and the signal level of the image signal of a different type from the two types of image signals used in the first subtractor, and a total of two outputs of these subtractors Two-dimensional mapping means for outputting, as the address information, an identification signal indicating whether the color of the image information is a chromatic color or an achromatic color and the type of the color when the color is chromatic, and a signal level of one of the three types of image signals. In the case where the color of the image information is an achromatic color by using the multivalued means for multivalued, and the data multivalued by this multivalued means and the identification signal output from the two-dimensional mapping means as the address information. It distinguishes between lightness and black and white, and in the case of chromatic colors, the color type and It is obtained by a color identifying means for outputting a color identification signal including a brightness or identification information of black and white if the achromatic.

In this color identification device, the outputs of the two subtractors are used to identify whether the color of the image information is a chromatic color or an achromatic color, and the type of the color in the case of the chromatic color, by using the two-dimensional mapping means, and the multi-value quantization means. And the output of the two-dimensional mapping means are used to identify the lightness or black and white when the color of the image information is achromatic. If this apparatus is provided with a resolution correction means for performing resolution correction on the image signal before being multi-valued, the generation of color ghost can be prevented more effectively. For example, a Laplacian filter is effective as the resolution correcting means.

[Examples] The present invention will be described in detail below with reference to Examples.

[First Embodiment] FIG. 1 shows a structure of a color identification apparatus according to a first embodiment of the present invention. This device has three types of image signals obtained by reading the same part of the original from a reading device (not shown).
1R, 1G and 1B are supplied. Here, the image signal 1R is a digital signal in which the signal level of the red wavelength component is represented by 8 bits, and the image signal 1G is a digital signal in which the signal level of the green wavelength component is represented by 8 bits. Image signal 1B
Is a digital signal in which the signal level of the blue wavelength component is represented by 8 bits.

In this embodiment, the image signal 1R is supplied to the first subtractor 2-1 together with the image signal 1G, where the difference between the two signal levels is obtained. Further, the image signal 1R is supplied to the second subtractor 2-2 together with the other image signal 1B, and the difference in signal level is obtained here as well. These difference signals 3-1 and 3-2 after the subtraction are 9-bit signals, respectively, and are supplied to the mapping ROM (read only memory) 4.

The significance of mipping using the difference signals 3-1 and 3-2 will be described. In a gray-balanced color original reading apparatus, when an achromatic portion of an original, that is, a generally gray portion is read, the above-mentioned three types of image signals are read.
The signal levels of 1R, 1G and 1B are the same. When the saturation of the read portion of the document increases, the difference in the signal levels of these image signals 1R, 1G, and 1B increases with each other. If the hue of the read portion of the document changes, the relative levels of the image signals 1R, 1G, and 1B change accordingly.

Therefore, in this color identification apparatus, first, the signal level is subtracted for the pair of image signals after the three-color separation, and the two-dimensional mapping is performed by the mipping ROM 4 by the output of these. Then, the achromatic color is mapped near the origin of the map as shown in FIG. Further, as the saturation increases, as described above, it is mapped at a position farther from the origin. The mapped position in the case of a chromatic color will differ depending on the hue.

Therefore, in this color identification device, two-dimensional mapping is performed, and (i) whether or not the image information of the document is an achromatic color is identified from the result, and (ii) if it is a chromatic color, the type of that color Identify {Yellow (Y), Magenta (M), Cyan (C), Red, Blue and Green}. At this stage, it is not possible to identify even the achromatic brightness. The above result is output from the mapping ROM 4 as a 3-bit identification signal 5.

This identification signal 5 is supplied to the monochrome identification ROM 6 together with the image signal 1G. The black-and-white identification ROM 6 identifies the degree of whiteness or blackness when the identification signal 5 identifies a portion of the document as an achromatic color. Then, it is combined with the output of the two-dimensional mapping and output as the 3-bit color identification signal 7.

The following Table 1 shows the average value of each signal level (256 steps from “0” to “255”) of the image signals 1R, 1G, and 1B after the three-color separation output by the color reading device. Subtractor 2
-1 and the subtraction result by the second subtractor 2-2.

For example, in the uppermost column of this table, the signal level of image signal 1R is
It is 58.3. Assuming that the signal level of the image signal 1G is 1.2 and the signal level of the image signal 1B is 26.2, the difference signal 3-1 as the output of the first subtractor 2-1 becomes 57.0, and the second
The output of the subtractor 2-2 becomes 32.1. In this case, the two-dimensional mapping is as shown by the point R1 in FIG. 2, and the identification result is red.

In the same way, the image information on the manuscript is sequentially determined,
Each of yellow, magenta, cyan, red, blue, green, and achromatic colors (white and black) is determined. The setting of the boundary line of each color in the two-dimensional mapping shown in FIG. 2, that is, the determination of the contents to be written in the mapping ROM 4 is performed using, for example, the results of reading various color samples with a color reading device. Assuming that "0" is the black signal level and "255" is the white signal level, the threshold (threshold level) for distinguishing between white and black when reading achromatic colors in the brightness identification ROM6 is, for example, about "160". Is appropriate.

In the first embodiment, the image signal 1R is generated by the first subtractor 2-1.
From the image signal 1G, and the second subtractor 2-2 subtracts the image signal
Although the image signal 1B is subtracted from 1R, it goes without saying that the signal level difference may be obtained in different combinations.
In this case, the contents of the mapping ROM 4 will be changed accordingly. Further, although 6 chromatic colors are identified in the embodiment, when it is necessary to identify 7 or more chromatic colors, the mapping table shown in FIG. 2 is further subdivided in correspondence with these colors. You can change it.

[Second Embodiment] FIG. 3 shows a color identification apparatus according to a second embodiment of the present invention. In FIG. 3, the same parts as those in FIG. 1 are designated by the same reference numerals, and the description thereof will be appropriately omitted.

By the way, in a normal color reading apparatus, image information on a document is optically projected on an image sensor to identify a color on a pixel-by-pixel basis. In this process, the resolution is deteriorated, and as a result of the color identification for each pixel, the image is missing in the thin line portion, or the image is crushed in the complicated combination and color pattern portion.

As a method of eliminating such a problem caused by resolution deterioration, resolution correction is performed on each image signal separated into three colors. However, when such a method is adopted, another color that is not on the original document may occur as a ghost between the actual color existing on the original document and the color due to signal processing accompanying resolution correction. . Further, since each image signal is independently subjected to resolution correction, it is necessary to prepare a total of three systems of resolution correction means such as a Laplacian filter for each color, which causes a problem that the color identification device itself becomes expensive.

The color identification device of the second embodiment is a device that can solve such a problem.

In this device, the image signal 1G is supplied to the first subtractor 2-1 and also supplied to the Laplacian filter 11 to correct the resolution. The corrected image signal 12G is supplied to the black-and-white identification ROM 6 and compared with a predetermined threshold value to determine whether the achromatic color is white or black.

FIG. 4 shows a specific example of the Laplacian filter used in this embodiment. The image signal 1G is latched by the D flip-flop 21. The output of the D flip-flop 21 is input to the other D flip-flop 22 and is also input to the first line memory 23, and is delayed by one line in the read scan. The delayed image signal is latched by the D flip-flop 24, and is input to the next D flip-flop 25, and is also input to the second line memory 26 and further delayed by one line in the reading scan. The output of the second line memory 26 is sequentially shifted by the two-stage D flip-flops 27 and 28, and the result is input to the adder 29. The output of the D flip-flop 22, the output of the other D flip-flop 31 provided in the subsequent stage of the D flip-flop 25, and the output of the second line memory 26 are also input to the adder 29.

In this way, the adder 29 becomes the D flip-flop 25.
The image signals for the peripheral four pixels centering on the image signal (image signal for the pixel of interest) 32 latched at are simultaneously received, and a predetermined addition operation is performed using them. Adder
The addition result 33 of 29 becomes the address information of the ROM 34 together with the image signal 32 latched in the D flip-flop 25, and the resolution-corrected image signal 12G of the target pixel is obtained.

In this way, the Laplacian filter 11 delays the image signal in the process of extracting the peripheral pixels. So for mapping
The output of the ROM 4 is also input to the delay circuit 36, and the identification signal 5 is delayed according to the delay amount of the image signal 12G. The black-and-white identification ROM 6 has an identification signal 37 supplied from the delay circuit 36.
Represents an achromatic color, this is compared with a predetermined threshold value (for example, about "160" as in the first embodiment) to identify black and white.

In the color identification apparatus according to the second embodiment of the present invention described above, the resolution correction is performed on the image signal 1G, but instead of this, the resolution is corrected for one of the other two types of image signals 1R and 1B. Correction may be performed. Further, resolution correction can be performed by a method other than the Laplacian filter method, and the same effect can be obtained by, for example, the adaptive binarization method.

In the second embodiment, the resolution of the achromatic color is corrected just before the white and black are discriminated. Therefore, resolution correction for achromatic colors can be performed without impairing the stability of color identification for chromatic colors. On the other hand, in the color identification apparatus according to the second embodiment, the resolution is not corrected in the chromatic color portion. However, in a normal document, fine parts such as fine lines are often drawn in black, so there is almost no problem due to this.

[Invention and Effect] As described above, according to the present invention, the two-dimensional mapping means uses the two difference signals resulting from different combinations of the three types of image signals separated into three colors as the address information.
Whether the color of the image information is a chromatic color or an achromatic color, and the type of color in the case of a chromatic color are identified, and the color identification means is used by using the identification result by the two-dimensional mapping means and one of the three types of image signals. Since the brightness or the black and white is discriminated in the case where the color of the image information is an achromatic color, both the two-dimensional mapping means and the color discrimination means output the discrimination signal or the color discrimination signal using the two signals as the address information. Because it is
The address space can be remarkably reduced as compared with a device that performs three-dimensional mapping to identify a color, and ROM can be easily configured, and the device can be simplified.

[Brief description of drawings]

FIG. 1 is a block diagram of a color identification apparatus according to the first embodiment of the present invention, FIG. 2 is an explanatory diagram showing a state of two-dimensional mapping, and FIG. 3 is a color identification according to the second embodiment of the present invention. FIG. 4 is a block diagram of the apparatus, and FIG. 4 is a block diagram of a Laplacian filter used in the second embodiment. 1 ... Image signal, 2-1 ... First subtractor, 2-2 ... Second subtractor, 4 ... Mapping ROM, 6 ... Monochrome identification ROM, 7 ... Color identification signal, 11 …… Laplacian filter.

Claims (3)

[Claims] 1. A color identification device for separating image information on an original into three colors and reading the image information, and identifying the color of the image information based on the three types of image signals obtained thereby. A first subtractor for subtracting the signal levels of arbitrary two types of image signals of the image signals, and the same image signal as one type of the two types of image signals used in the first subtractor A second subtractor for performing a subtraction between the signal level of the image signal and the signal level of the image signal of a different type from the two types of image signals used in the first subtractor, and a total of two outputs of these subtractors are addressed. As information, a two-dimensional mapping unit that outputs an identification signal indicating whether the color of the image information is a chromatic color or an achromatic color, and a color type in the case of a chromatic color, and one of the three types of image signals Multi-valued means for multi-valued levels and multi-valued means by this multi-valued means Data and the identification signal output from the two-dimensional mapping means as address information,
A color identifying unit that identifies the lightness or black and white when the color of the image information is an achromatic color, and outputs a color identification signal including the type of color when the color is chromatic and the lightness or black and white identification information when the color is achromatic. A color identification device comprising: 2. The color identification apparatus according to claim 1, further comprising resolution correction means for performing resolution correction on an image signal before being multivalued by the multivalued means. 3. The color identification apparatus according to claim 2, wherein the resolution correction means is a Laplacian filter.