JPH04111180A - Processing method for color image - Google Patents

Abstract

PURPOSE:To generate n-number of binary picture with one conversion processing while designating n-number of colors by denoting the expression element number of color image to (j) and the number of colors to (n) and utilizing an output density bit width (n), an independent (j) number of density conversion LUT, an input AND means, and a gradation picture storage means. CONSTITUTION:A color image data 101 is composed of gradation images 101R, 101G, and 101B by three elements expressing the color image to respectively turned to be the input of the gradation conversion LUT 102R, 102G, and 102B for each element. Density conversion results 103R, 103G, 103B for each element are turned to be the input of a 3 input AND means 104, and an output 105 of the 3 input AND means 104 is stored in a gradation picture storage means 106. By setting the conversion function to the density conversion LUT 102R, 102G, and 102B, the picture data overlapping the binary picture generated by designating the plural colors to the gradation picture storage means 106 can be obtained. Thus, the high-speed and simple processing can be performed when dealing with the processing object where the plural colors are combined.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Field of Application] The present invention is based on the following methods: 1. pixels that can be considered to have a desired color from a digitized color image; Regarding a method of generating a binary image using color image processing, in which a predetermined arithmetic process is performed on a binary image generated with other pixels set to O to obtain a desired result, more specifically, a method for generating a binary image by specifying multiple colors and each The present invention relates to a method for obtaining a plurality of binary images corresponding to colors.

[Conventional technology]

In many color image processing devices, a color image is expressed by a combination of shading images for each of the three elements of RGB corresponding to the three primary colors of light. In other words, a digitized color image is represented by three digitized grayscale images.

In order to extract pixels of a specified color from a color image, first, the range of values of each element of RGB that can be considered to be a specified color is determined. Compare the value of each RGB element of a pixel in a color image with the range of values of each RGB element that can be considered as the specified color, and if all three elements are within the range, the pixel is the specified color. It is determined that

FIG. 5 is a diagram showing an example of a method for performing this processing at high speed, and is an explanatory diagram of a method using a look-up table (LUT), which is an image density conversion method. In this method, an LUT is created in advance that contains a density conversion function representing the relationship between input density and output density, and the output density value is immediately read out from the input density value, thereby speeding up the image density conversion process. This method is widely used in image processing and other applications because it is easy to implement in hardware and perform real-time processing. Prepare this LUT for each RGB element of a color image, and for each element's LUT, set the output density to 1 within the range of input density values that can be considered to be the specified color, and set the output density to 0 for other input density values. shall be. These three LUs
By converting the density of a color image using T and calculating the logical product of the three output densities, a binary image in which the pixel of the designated color is 1 and the other pixels are O is obtained.

[Problem to be solved by the invention]

However, in the conventional method described above, a binary image can be generated by specifying a color from a color image at high speed using an LUT, but only a binary image with one color specified can be obtained in one process. I can't.

Therefore, in order to specify multiple colors for one color image and process each pixel of each color, or to perform processing using a combination of multiple colors, such as calculating the sum of the areas of red and blue parts, for example. In this case, it is necessary to specify one color of each color and repeat the process of generating a binary image for the desired number of colors.

An object of the present invention is to specify a plurality of colors and generate a binary image for each color in which the pixels considered to be of that color are 1 and the other pixels are 0. An object of the present invention is to provide means for generating binary images corresponding to each of a plurality of colors specified in a degree conversion process.

[Means to solve the problem]

In order to achieve the above object, the present invention uses j independent density conversion LUTs whose output density bit width is n, where j is the number of elements expressing a color image and n is the number of colors handled simultaneously. , j inputs and one output have a bit width of n, and a grayscale image storage means in which the bit width of one pixel is n. The obtained color image data specifies up to n colors and generates up to n binary images in one conversion process.

[Effect]

In order to realize this conversion operation, "density conversion LUT
The density conversion function to be set is generated by the following procedure.

A specified color is specified when the input density values of all elements for expressing a color image are within the range expressed by the upper and lower limits of the values of each element of the color specified for generating a binary image. It is determined that To set the conversion function for the k-th color, set the 2-th bit of the output density value to 1 within the range of the corresponding element of the color specified for the density conversion LUT of a certain element, and set the output density for other input densities. By setting the 2nd bit of the value to O, 1 or 0 is set for the 2nd bit of the density conversion LUT of all elements.

Conversion functions are set for n colors using the above steps.
In each density conversion LUT, a conversion function is set for all bits. When the j-input logical product is calculated from the output density values from these three density conversion LUTs, only the bit that is 1 for all of the j output density values becomes 1, and the other bits become 0.

That is, if the value of each element of the color image data is within the specified color range for all elements, it is j, otherwise it is O, and the color image data is binarized. At this time, each bit is processed simultaneously while remaining independent, and if the k-th bit is 1, it means that the color image data at this time is determined to be of the i-th color.

The grayscale image storage means stores n images of 2 which are the results of binary image generation.
Value images are superimposed and stored as one image.

If we focus only on the bit number of the grayscale image storage means, it is a binary image obtained by specifying the th color.

〔Example〕

Hereinafter, one embodiment of the present invention will be described using the drawings. In this example, the color image is R corresponding to the three primary colors of light.
Although the case is shown in which it is expressed by three elements of GB, it can be realized by the same method even when using other expression methods.

FIG. 1 is a block diagram when binary images corresponding to a plurality of specified colors are simultaneously generated from a color image in a single conversion process according to the present invention. The color image data 101 obtained by reading out the color image input means or the color image storage means includes grayscale images 101R, 101G, and 10 for each of the three elements expressing the color image.
1B-C, and each element serves as an input to density conversion LUTs 102R, 102G, and 102B.

Density conversion results for each element 103R1103G,
106B becomes the input to the 3-man logical product means 1040, and 3
The output 105 of the manual logical product means 104 is stored in the grayscale image storage means 106.

By setting the conversion functions according to the above-described procedure in the density conversion LUTs 102R, 102G, and 102B, image data in which binary images generated by specifying a plurality of colors are superimposed in the grayscale image storage means 106 can be obtained. In the following explanation, for simplicity, the horizontal axis of the density conversion LUT is the input density,
The vertical axis is expressed as a 4-bit pit string.

FIG. 2 shows an example of setting a conversion function for the color number O. Specified color y' - ta%
';! (r OSg O, b O) The width of the value determined to be the color with number 0 is (rWo, g
wQ, bwo). In bit 0 of the density conversion LUT 201 for element R of the color image, the input density (rQ
±rWo/2) range is set to 1, and other input concentrations are set to 0. Density conversion LUT20 for elements G and B
Similarly, in 2.203, each (go+gWO
, /2), (bo+:bwo/2), and 0 for other values.

FIG. 3 shows an example of three density conversion LUTs when all four colors are specified.

In the figure, the hatched area represents 1, and the plain area represents O. When the color data of a certain pixel in a color image is (r3, g3, b3), the values read from each density conversion LUT 301, 302, 303 are (1oo1), (0001), (0101), respectively.
Therefore, from the logical product (0001) of the three values, the number 0
Figure 4 shows the grayscale image storage means 1.
06 (FIG. 1) is a schematic diagram when processing results are stored.

One pixel 411 of the grayscale image has a value expressed by 4 bits, and a collection of pixels for one screen constitutes the grayscale image 410. Focusing on each bit of the grayscale image 410, there are four 2
It can be seen as a superposition of the bamboo images 400 to 403. In the process of converting a color image according to the present invention, the output 105 (FIG. 1) of the three-manual logical product means is obtained in the form of one pixel 411 of a grayscale image, and when the processing for one screen is completed, the grayscale image 410 is It takes shape. The binarized converted images for each designated color for the color image are images 400 to 403 cut out for each focus of the grayscale image 410.

All bits of pixels that are not of any of the four designated colors are O. Processing that combines a plurality of colors can be easily performed from the bit pattern of the pixels of the grayscale image 410. For example, to find the total sum of pixels that are the color number O or the color number 2, the pixel value and the bit pattern (0101) can be logically summed to find the number of non-zero pixels.

〔Effect of the invention〕

As described above: According to the present invention, multiple colors can be specified.
Since binarization is performed independently for each color in one conversion process, high-speed and simple processing can be performed when processing a processing target that is a combination of a plurality of color cars. In addition, it is a combination of general methods used in conventional color image processing, such as a density conversion LUT for each element of a color image, a three-way logical product method, and an m-light image memory, and does not require any special hardware. Therefore, it can be easily implemented.

[Brief explanation of the drawing]

1 to 4 relate to the present invention; FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is an explanatory diagram of the procedure for creating a density conversion table according to the present invention, and FIG. FIG. 4 is an explanatory diagram of the principle of converting a color image into a plurality of binary images using the density conversion table according to the present invention. FIG. FIG. FIG. 5 is a block diagram illustrating conversion of a color image into a binary image according to a conventional technique. 102R1102G, 102B...Density conversion L
LTT. 104...3 manual logical product means, 106...
... Grayscale image storage means. Input product in 25th Figure 3, -, (oool) Figure 4

Claims (1)

[Claims] Predetermined arithmetic processing is performed on a binary image generated from a digitized color image by setting pixels that can be considered to have a desired color as 1 and other pixels as 0 to obtain a desired result. In color image processing to obtain images, each binary image generated by specifying multiple colors is selected individually or combined using logical operations, and predetermined arithmetic operations are performed on the binary images. When performing color image processing for multiple colors, if the number of elements expressing the color image is j, and the number of colors handled simultaneously is n, then A density conversion lookup table (hereinafter referred to as a density conversion LUT), a j-input logical product means in which the bit width of j inputs and one output is n, and a grayscale image storage means in which the bit width of one pixel is n are used. , (a) the colors specified for generating the binary image,
(b) the kth color of the colors specified to generate the binary image; To set the conversion function for color, in the density conversion LUT for each of the j elements, set the kth bit of the output density value to 1 within the range of the input density value, and set the output density value to 1 for other input density values. By setting the kth bit of the density value to 0, set the kth bit in the density conversion LUT for all elements, and follow the same procedure for the n colors specified to generate the binary image. By setting the conversion function, the j density conversion LUs
(c) The j elements of the color image data are subjected to density conversion using the j density conversion LUTs, and the j elements of the color image data subjected to the density conversion are calculating the logical product of the values of all the elements by the j input logical product means, and writing the logical product of the j elements of the density-converted color image data into the grayscale image storage means; (d) the grayscale image storage means; By reading only the k-th bit of the image data written in the image data, the pixel determined to be the k-th color out of the n colors specified for generating the binary image is 1, and the other pixels are Color image processing characterized by specifying n colors for a color image by obtaining a binary image in which pixels are 0, and generating a binary image with n screens in one conversion process. Method.