KR100408508B1 - Method and apparatus for processing color, signal using color difference plane separation - Google Patents

Abstract

PURPOSE: A method and an apparatus for processing color using color difference plane division are provided to simply realize selective color adjustment and color correction. CONSTITUTION: A coefficient table(11) divides a color difference plane defined by two color difference signals into a plurality of cells. The coefficient table obtains coefficients of transformation matrix for converting demodulated input color signals into output color signals. The coefficient table groups the cells having the similar coefficients in characteristic. An index address table(16) stores a coefficient designating index for designating addresses of the coefficients of the transformation matrix, and generates the coefficient designating index by using a plurality of upper bits of the two color difference signals. A matrix operating unit(17) operates the coefficients of the transformation matrix corresponding to the coefficient designating index with the input color difference signal. A range checking unit(18) generates the output of the matrix operating unit as the output color signal if each value of the color signal is within an allowable range. If out of the allowable range, the maximum value within the allowable range is generated as the output color signal.

Description

Method and apparatus for processing color, signal using color difference plane separation}

The present invention relates to a color processing method and apparatus for color correction and color correction of inaccurate and uncertain parts of color information of an original image due to a problem of a display device including a cathode ray tube (CRT) in a display device including a color television. In particular, the present invention relates to a color processing method and apparatus having a flexible and simple structure capable of independently correcting both color signal distortion factors and visually sensitive factors in a color display device.

In general, color display devices, especially color television, demodulate color signals produced according to broadcasting standards (NTSC, PAL) and reproduce them through a cathode ray tube (CRT). However, the received color information causes color distortion by various factors, one of which is the color signal processing in the TV receiver. That is, the color reproduction difference between the input color and the CRT output color, which are different from the broadcasting standard specification, in which the R, G, and B phosphors of the CRT are predetermined, and the color deviation of intermediate colors due to nonlinearity of the receiver circuit, etc. There is this. In addition, in an output device such as a color television, it is necessary to partially compensate for the lack of the primary colors or to visually process color due to differences in color viewpoints according to races.

As a representative method of the prior art to solve this problem, US Patent No. US Pat. No. 4,695,875 shows a number of color correction devices for correcting a received color signal according to a person's preference color by adjusting two reference phases of a chroma demodulator in the case of a TV. Although it was used, there were disadvantages in that the overall color distortion of the other colors was caused.

As another conventional technique related to the characterization of color display apparatuses, there are methods for modeling the relationship between input and output color values over the entire color space, and representative examples are a matrix method and a look-up table (LUT) using regression analysis. ) + Volume interpolation. Among them, LUT + volume interpolation measures a number of color values to obtain the input / output relationship of the system, and expresses the intermediate values using these measurement points and volume interpolation together. This method is relatively widely used in the printing field because it can obtain relatively accurate results compared to the existing method, and also enables local color adjustment for visual color processing. However, since volumetric interpolation is complicated, hardware (H / W) for real-time processing is difficult in display devices such as TVs, and it is not suitable for commercial applications because it is expensive. In addition, the regression analysis method is a method for simple matrixing of input / output relations based on a plurality of measurement points, and is widely used in the display field. However, since it is necessary to cover the entire color space with one matrix, it is not a big advantage in terms of color difference, and there are also disadvantages that local color tone, such as visual color tone, is impossible.

As a recent method to solve this problem, color display using regression analysis for each section by dividing five sections of flesh, gray, red, green, and blue There have been prior art attempts to characterize devices. However, by dividing the color into only five representative colors, a boundary line may appear between the divided regions, and it is difficult to adequately represent characteristics of various colors represented by color mixing in a color display device. In addition, it is also difficult to provide such a function when local color tone adjustment is required in a delicate color part, which is a mixed color of five representative colors.

As another prior art, US Patent No. US Pat. No. 4,989,080 shows a method of dividing six color regions bounded by six colors of RGB CMY on a chromaticity diagram and independently obtaining transfer functions in each color region. However, this method is also unsuitable for fine local color adjustment, and when more color division is performed, the color gamut is complicated and there is no countermeasure against saturation division.

Accordingly, an object of the present invention is to change colors in a three-dimensional color space into a region of colors on a two-dimensional color difference plane (eg, RY, BY) that is easy to classify and colorize hues and chroma. Color difference plane conversion method, color difference plane dividing method that divides two-dimensional color difference plane into a number of small areas, and color difference minimization method that defines the input / output relationship including the surrounding area adjacent to the corresponding area in each cell The present invention provides a color processing method and apparatus for performing color processing. Therefore, color display device characterization and color boundary can be solved.

In addition, another object of the present invention is to provide a color processing method and apparatus that is easy to directly local color correction and coefficient adjustment in color difference signals (R-Y, B-Y) using a plurality of divided small areas. Therefore, real time processing H / W can be implemented in a display device such as a TV.

In order to achieve the above object, the color processing method according to the present invention comprises: a) dividing a color difference plane defined by two color difference signals into a plurality of areas, and converting an input color signal received and demodulated by the display device into each color area as an output color signal; Obtaining coefficient values of the transformation matrix for transforming, and grouping and storing regions having similar characteristics of coefficient values for each partition region; b) if one division region on the color difference plane is designated by two color difference signals in the input color signal, obtaining coefficient values of the transformation matrix obtained in step a) for the designated division region; c) matrix computing the coefficient values of the transform matrix obtained in step b) and the input color signal; And d) checking whether each value of the color signal exists within an allowable range as a result of the matrix operation in step c) and outputting the maximum value within the allowable range when exceeding the allowable range.

In order to achieve the above objects, the color processing apparatus according to the present invention divides a color difference plane defined by two color difference signals into a plurality of regions, and converts an input color signal received and demodulated by a display device into an output color signal for each divided region. A coefficient table for obtaining a coefficient value of a transform matrix to be grouped and grouping and storing an area having similar characteristics of coefficient values for each partition area; An index address table for storing a coefficient designation index for designating an address of a coefficient value of a transform matrix stored in the coefficient table, and generating a coefficient designation index using a plurality of upper bits of two color difference signals in the input color signal; A matrix operation unit performing matrix operation on the coefficient table of the transform matrix corresponding to the coefficient designation index generated from the index address table in the coefficient table with the input color signal; And a range checker for generating an output of the matrix calculator when the value of the color signal is within an allowable range, and generating a maximum value within the allowable range as the output color signal when the matrix operation is out of the allowable range. It features.

1 is a schematic configuration diagram of a display apparatus to which a color processing apparatus according to the present invention is applied;

Figure 2 is a detailed configuration example showing an embodiment of a color processing apparatus according to the present invention.

3 is a detailed structural diagram example showing another embodiment of a color processing apparatus according to the present invention;

4 is a color expression example in a color difference plane.

5 is an example of interest color definition and 20 * 20 area division in a color difference plane.

6 is an example of local color correction in a color difference plane.

7 is a configuration diagram of an index address table and a coefficient table.

Hereinafter, with reference to the accompanying drawings will be described in detail a preferred embodiment of the present invention.

1 is a schematic configuration diagram of a display apparatus to which a color processing apparatus according to the present invention is applied. That is, for example, when the present invention is applied to a TV receiver, a modulated color signal received by the TV receiver is output to the demodulator 2 by outputting a video signal through the tuner & IF stage 1, and demodulator 2. One luminance signal (Yin) and two color difference signals (RYin, BYin) are generated by using the reference phase signal at. These signals are input to the color processor 3 proposed in the present invention, and different color conversions are performed for each area position in the color difference plane, and the output values (Yo, RYo, BYo) are Y, RY, BY-to-RGB. It is input to the converter 4 and is converted into Ro, Go, and Bo, which are RGB signals, which are signals required for driving the CRT 5, and are represented in the CRT 5.

One embodiment of the color processor proposed in the present invention is an apparatus for converting an input color signal into a desired output color signal by using a color difference signal and a luminance signal together, and will be described with reference to FIG. 2.

In FIG. 2, among the luminance signal Yin and the color difference signals RYin and BYin, which have passed through the demodulator (2 in FIG. 1), the luminance signal Yin that is not related to color specification is a 3 * 3 matrix calculator 14. ) And two chrominance signals related to color conversion are divided into two paths, one of which is divided into chrominance areas obtained by inputting only a few upper bits (MSB) of each signal into the index address table 11. Used as an index signal specifying one of these.

The configuration of the index address table 11 is two-dimensional and is divided into an area of 16 * 16, for example, each of which has coefficient designation values capable of addressing coefficient sets stored in the coefficient table 13. As a memory, a color difference signal composed of MSBs is input, and coefficient designation indexes I1 and 12 that can designate one coefficient set in the coefficient table 13 are output.

The coefficient table 13 is a memory that stores a plurality of matrix sets in which nine coefficients are configured as one matrix set. The coefficient table 13 receives a coefficient designation index 12 and converts nine coefficients stored at a corresponding address into a 3 * 3 matrix calculator. Output to (14).

The 3 * 3 matrix calculating section 14 receives nine coefficients from the coefficient table 13, while inputting two color difference signals RYin and BYin and a luminance signal Yin, which are output signals of the demodulation section 2, respectively. And a 3 * 3 matrix operation, which outputs the result to the range checker 15.

The range checker 15 performs a function of changing a value exceeding an allowable range of the luminance and chrominance signal to a maximum value within the allowable range, and the result is Y, RY, BY-to-RGB conversion. Output to negative (4 in FIG. 1).

Another embodiment of the color processor proposed by the present invention is an apparatus for converting an input color signal into a desired output color signal using only a color difference signal, which will be described with reference to FIG. 3.

In Fig. 3, among the luminance signal Yin and the color difference signals RYin and BYin, which have passed through the demodulator (2 in Fig. 1), the luminance signal Yin not related to the color representation is output without any further processing ( Ie Yo = Yin). The two color difference signals related to the color conversion are divided into two paths, one of which is that only the upper few bits (MSB) of each color difference signal are input to the index address table (11) to divide one of the divided color difference areas. Used as an index signal to specify.

The configuration of the index address table 11 is two-dimensional, for example, divided into 16 * 16 areas, each of which stores coefficient designation values capable of addressing coefficient sets stored in the coefficient table 16. As a memory, a color difference signal composed of MSBs (when the area is 16 * 16; 4 bits) is input and outputs coefficient designation indexes I2 and 12 that can designate one coefficient set in the coefficient table 16. The coefficient designation index 12 has a different number of bits according to the size of the coefficient table 16 used. For example, when 128 matrix coefficients are used in the coefficient table 16, 7 bits are assigned to the coefficient designation index.

The coefficient table 16 is, for example, a memory that stores a plurality of matrix sets in which four coefficients are formed as one matrix set. The coefficient table 16 receives four coefficient storage signals from the index address table 11 and stores four coefficient sets. The coefficients are output to the 2 * 2 matrix calculator 17.

The 2 * 2 matrix calculator 17 receives four coefficients from the coefficient table 16, and receives two color difference signals RYin and BYin, which are output signals of the demodulator 2, to perform 2 * 2 matrix operations. As a matrix operator for performing the operation, the calculation result is output to the range checker 18.

The range check unit 18 performs a function of changing a value exceeding an allowable range of the color difference signal to a maximum value within the allowable range, and the result is Y, RY, BY-to-RGB conversion unit (Fig. Output to 1) 4).

4 shows an example of a typical color difference plane using color difference signals of RY (R-Y) and BY (B-Y). Any one point in the chrominance plane represents a color with a particular hue and saturation, and all colors in the three-dimensional color space can be described in the two-dimensional chrominance plane. The technique can be easily expressed in color and saturation as shown in Equations 1 and 2 below.

All colors can be distinguished by different colors (angle; B-Y axis = 0 degrees) and saturation (length; gray point = 0) on the chrominance plane.

FIG. 5 is a two-dimensional color difference plane divided by 20 * 20. The divided cells have different definitions of color and saturation, and flesh color, wood color, and sky color, which are of interest colors, also occupy areas different from each other. Doing. Therefore, the relationship between the input color and the output color per area can be represented more precisely with a simpler formula than the definition of the input / output relationship for the entire color space.

For example, the process of obtaining the coefficients for each divided small region is described as follows.

Step A1) A plurality of different colors covering the color difference plane are input, and the corresponding colors are measured by the color coordinate values such as CIExyY in the CRT.

Step A2) All measured CIExyY values are converted into Y, RY and BY color values based on the NTSC formula as shown in Equation 3 below.

Step A3) By dividing the color difference plane into the appropriate scale (16 * 16 or 32 * 32) and projecting the converted RYr and BYr onto the divided color difference plane, a set of measurement values belonging to each area can be obtained. .

Step A4) Input the input color signals Yi, RYi, BYi to the actual CRT using error minimization methods such as Yi, RYi, BYi and measured value XYZm of the input signal in each area, and the value XYZc information to be originally expressed and regression analysis. The coefficient value of the conversion matrix M, which is color-converted into the color signals Yc, RYc, and BYc, to be obtained, can be obtained as in Equation 4 below.

Step A5) The cell grouping is repeatedly performed for all the regions in which the regions having similar characteristics of coefficient values among the coefficient values of the region-specific transformation matrix in the color difference plane obtained in Step A4 are collected and repeated for all regions. Minimize the number of matrices.

Referring to FIG. 5 as an example, the four regions designated as A are equivalent to one region because they use the same conversion matrix. In the case of the notation B, three regions are one region, and in the case of the notation C, region grouping occurs. Shows no.

Referring to FIG. 6, a process of obtaining coefficient values for each segmented area by using local color tone / color conversion as an example is as follows.

Step B1) The color tone specification is defined in the color difference plane. When tinting hue and saturation in any direction on the chrominance plane, for example, the angle of the color to be shifted (cLCH [2]), the angle of the color when shifted (c2LCH [2]), and both The angle of the interface color (uLCH [2], dLCH [2]), the amount of saturation to be shifted (cLCH [1]), the amount of saturation when moved (c2LCH [1]), and the maximum / minimum saturation Sizes uLCH [1] and dLCH [1] are defined respectively. In addition, DELTA C, DELTA H, and DELTA T are defined as follows.

ΔC = c2LCH [1] -cLCH [1],

ΔH = c 2 LCH [2] -cLCH [2],

ΔT = ΔH and ΔC

Step B2) If the color difference plane is cell-divided into appropriate scales (16 * 16 or 32 * 32) and the two-dimensional ranges of both boundary angles and maximum / minimum saturation of the colors participating in the hue definition are projected onto the color difference plane, It is possible to define the partition in the chrominance plane that participates in the adjustment.

Step B3) The definition of the change amount of hue / saturation at all positions within the selected entire area may be triangular, trapezoidal, or other functions, and may be expressed as in Equations 5 and 6 below.

Step B4) The coefficients of the transformation matrix in each selected partition can be obtained by error minimization method such as regression analysis as a transfer function between the original color signal and the color signal changed through step B3 (similar to step A4).

If all the coefficient values of the conversion matrix for each color difference plane region divided through steps A1 to A5 and B1 to B4 are obtained, the coefficient values for each region of the color difference plane are the coefficient table 13, for example. The coefficient designation index values, which are stored in the coefficient table 13 and which designate the coefficient address stored in the coefficient table 13, are stored in the index address table 11. On the other hand, taking FIG. 3 as an example, the coefficient values of the region-specific transformation matrix of the color difference plane are stored in the coefficient table 16, and the coefficient designation index values for designating the coefficient addresses stored in the coefficient table 16 are index address table 11 Are stored in. 7 is a diagram showing the mapping relationship between the index address table 11 and the coefficient tables 13 and 16.

The present invention can be implemented by using a simple circuit in all color display devices that perform digital signal processing, namely, color display devices including color televisions and color input / output devices. Easily applicable to processing. The present invention can also be used for general purpose IC chip formation.

As described above, the present invention divides into a plurality of regions using a color difference plane, and calculates input / output relations in each divided region as a simple linear matrix, such as color adjustment, green / sea emphasis, etc. in addition to accuracy and device characterization. The optional tint / color correction is simple and has a flexible structure that can be changed.

In addition, since the apparatus required for implementing the present invention is not complicated, and the pixel processing time is simple and the calculation time is small, it can be used for real time moving picture processing.

In addition, as a structure that is not restricted by the device used, it can be widely used in all display devices that perform digital signal processing such as TV, and other input / output devices such as cameras and printers, and can also be used in general-purpose IC chips. .

Claims (2)

a) The color difference plane defined by the two color difference signals is divided into a plurality of areas, and the coefficient value of the conversion matrix for converting the input color signal received and demodulated by the display device into the output color signal for each divided area is obtained, and for each divided area. Grouping and storing regions having similar characteristics of the coefficient values; b) if one division region on the color difference plane is designated by two color difference signals in the input color signal, obtaining coefficient values of the transformation matrix obtained in step a) for the designated division region; c) matrix computing the coefficient values of the transform matrix obtained in step b) and the input color signal; And and d) checking whether each value of the color signal exists within the allowable range and outputting the maximum value within the allowable range when the result of the matrix operation in step c) exceeds the allowable range. The color difference plane defined by the two color difference signals is divided into a plurality of areas, and for each divided area, a coefficient value of a conversion matrix for converting an input color signal received and demodulated by the display device into an output color signal is obtained, A coefficient table for grouping and storing regions having similar characteristics of values; An index address table for storing a coefficient designation index for designating an address of a coefficient value of a transform matrix stored in the coefficient table, and generating a coefficient designation index using a plurality of upper bits of two color difference signals in the input color signal; A matrix operation unit performing matrix operation on the coefficient table of the transform matrix corresponding to the coefficient designation index generated from the index address table in the coefficient table with the input color signal; And And a range checker for generating an output of the matrix calculator when the value of the color signal is within an allowable range, and generating a maximum value within the allowable range as the output color signal when the value is outside the allowable range. Color processing apparatus.