JPH0758969A - Color image converting method - Google Patents

Abstract

PURPOSE:To obtain a color image with satisfactory color reproducibility and high picture quality by obtaining a signal for output to output a color near to the one corresponding to a chrominance signal. CONSTITUTION:A luminance signal and a color difference signal, after being converted into three ink density signals as output signals by a conversion table as a color conversion circuit, are supplied to a color printer as an output means. In such a case, a pair of standardized chrominance signals in accordance with a luminance/color difference signal Y are found, and the point of it is set as Y'. After that, a point in a curved surface body i.e., a pair (X1, X2, X3) which minimizes differential distance in the pair (X1, X2, X3) of standardized chrominance signals in accordance with a pair of density of outputtable ink is found, and the conversion value of a pair of luminance/color difference signal in accordance with the pair of standardized chrominance signals is set equal to the conversion value of the signal Y. In this way, when the luminance/color difference signal without corresponding to an outputtable color is inputted, it is assumed as the one equivalent to the input of a signal with highest likelihood.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a color image conversion method for a color image output device for printing or displaying a color image.

[0002]

2. Description of the Related Art In recent years, with the development of color printer technology, research and development for practical use of color copying machines have been actively pursued. The color copying machine photoelectrically converts the original image information and forms a color image by electrical processing,
There are two types of systems that form a color image by a chemical process as in the conventional copying machine. Compared with the latter method, the former method has drawbacks such as a lot of noise and low resolution at present, but it has the advantage that it does not require a chemical process and can perform various processing in image formation. Has received a great deal of attention.

As an example of a color copying apparatus that forms a color image by electrical processing, original image information on the surface of an original or the like is detected as, for example, three color signals, and this is converted into a digital signal by an A / D conversion circuit. Then, it is separated and converted into a luminance signal and two color difference signals with low correlation, and R
There is known a method in which a color image is printed by supplying a color printer as an output unit after converting an output signal suitable for an output form, that is, an ink density signal by a conversion table using the OM.

Here, the conversion table converts the luminance / color difference signal into an ink density signal according to the characteristics of the ink used in the color printer, but this conversion is generally a complicated calculation. Therefore, the A / D conversion up to the previous stage,
While a shading correction, a gamma correction and a process such as luminance and color difference signal separation are executed by a relatively simple arithmetic circuit, a table is used for this conversion. The structure of this conversion table is such that a certain set of luminance / color difference signals corresponds to a set of ink densities expressing the colors represented by the set of luminance / color difference signals, which can be obtained by numerical calculation. it can. However, in reality
The set of color difference signals does not represent the actual color, and
There are some colors that represent actual colors but cannot be represented by the ink used in the color printer. For such a combination of a luminance / color difference signal that does not correspond to a color that can be displayed by a color printer, the corresponding ink density cannot be determined by numerical calculation. However, in reality, due to noise mixed in the image input section and input of colors that cannot be expressed with ink used in color printers, the
There is a high possibility that a color difference signal will be input, and in such a case the color printer will not respond at all, which causes deterioration of image quality.

[0005]

SUMMARY OF THE INVENTION It is an object of the present invention to add a color signal to a color signal that does not correspond to a color that can be output due to noise or color signal input of a color that cannot be output. Correspondingly it is possible to perform appropriate image output,
An object of the present invention is to provide a color image conversion method capable of obtaining a high quality color output image.

[0006]

According to the present invention, a first step of inputting a plurality of detected color signals and a color signal input in the first step are adapted to an output form of a color image output device. And a second step of converting the color signal into the output signal, the second step of obtaining a predetermined color signal corresponding to the color signal input in the first step. And the color signal input in the first step does not correspond to a color that can be output by the color image output device, a predetermined color that can be output by the color image output device. In the fourth color step, and a fourth step of finding a predetermined color signal that minimizes the error distance E from the predetermined color signal obtained in the third step. The specified color signal 5 to the output signal of the output signal to respond to said first input color signal in step
It is characterized by including the step and.

[0007]

According to the present invention, when an input color signal (for example, a set of luminance and color difference signals) is converted into an output signal (ink density signal in the case of a color printer) suitable for an output form of a color printer or the like. , It is possible to obtain an appropriate output signal for every set of color signals that can be input, including those due to the influence of noise and the like.

That is, according to the present invention, for the color signals that do not correspond to the colors that can be output, the predetermined color signals (x1, x2, x3) that correspond to the colors that can be output by the output device.
) Corresponds to a predetermined color signal (x1, x2, x3) that minimizes the error distance E with respect to the predetermined color signal (y1, y2, y3) corresponding to the input color signal. Since the same output signal value as the output signal value is used as the conversion value for the input color signal, it is possible to output a color close to the original signal while maintaining vividness.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows an embodiment in which the present invention is applied to a color copying machine. In the figure, an image input unit 1 detects original image information as a plurality of color signals 2, and is configured as shown in FIG. 2, for example. That is,
A white light source 22 illuminates a document surface 21 as a plurality of objects,
The image on the document surface 21 is formed on the light receiving surface of the photodetector array 24 by the lens 23. On the front surface of the photodetector array 24, filters 25 of three types of colors having different spectral transmission characteristics are arranged corresponding to three consecutive photodetector arrays 24. Then, three consecutive sets of the photodetector array 24 form one image element, and photoelectric conversion outputs of these three sets are obtained as three color signals 2 having color information of pixels.

The color signal 2 thus obtained by the image input unit 1
Is subjected to parallel-serial conversion in pixel units and becomes a time-series signal. These color signals 2 are digitized by the A / D converter 3, and the digitized color signals 4 become standardized color signals 6 by the shading correction circuit 5. The shading correction circuit 5 has a correction data table of sensitivity and offset for each photodetector. Through this table, the color signal 4 is corrected for variations of each pixel of the photodetector, and the input pixels are white and black. In this case, the values are standardized so as to take the maximum value "1" and the minimum value "0", respectively, and output. The standardized color signal 6 is separated by a brightness / color difference signal separation circuit (Y / C separation circuit) 7 into one brightness signal 8 and two color difference signals 9a and 9b. The three input signals (normalized color signals) 6 of the Y / C separation circuit 7 are x1, x2, x3.
If the output signals 8, 9a, 9b are y1, y2, y3,
The Y / C separation circuit 7 performs matrix calculation as shown in the following equation.

[0011]

[Equation 1] Among the outputs of the Y / C separation circuit 7, the luminance signal 8 is a signal representing the brightness of the pixel, the color difference signals 9a and 9b are the signals representing the hue of the pixel, and when the input pixel is achromatic, the color difference signal 9a, Both 9b are 0.

The luminance signal 8 is further subjected to non-linear conversion by the gamma correction circuit 10. The input / output characteristics of the gamma correction circuit 10 are shown in FIG. As is well known, this gamma correction is intended to match that the human visual characteristic has a higher gradation discrimination ability in a dark portion. The scales of the color difference signals 9a and 9b output from the Y / C separation circuit 7 and the luminance signal 8'obtained via the gamma correction circuit 10 are close to the characteristics of human gradation discrimination.

Then, the luminance signal 8'and the color difference signal 9
A and 9b are converted into three ink density signals 12 as output signals by a conversion table 11 as a color conversion circuit, and then supplied to a color printer 13 as output means. The color printer 13 controls the ink density of each color by the ink density signal 12 and prints a color image. The conversion table 11 outputs an ink density for synthesizing a color corresponding to an input luminance / color difference signal, and has a three-dimensional structure.

The conversion table 11 is constructed as follows. That is, a color signal in the case of giving a set of arbitrary densities of three color inks used in the color printer 13 and inputting a color output therefrom can be obtained by calculation from the characteristics of the ink and the color filter. A set of luminance and color difference signals can be obtained. Here, the conversion table 1 for the original ink density set for this brightness / color difference signal set
By setting a conversion value of 1 (output signal value), correct color reproduction is possible.

However, as shown in FIG. 4, the set of luminance / color difference signal pairs corresponding to all the printable ink density pairs does not cover the entire possible luminance / color difference space. That is, in FIG. 4, 1 is the luminance signal axis,
Symbols y and q indicate two color difference signal axes, and a curved surface body 41 in the figure indicates a region of a set of luminance / color difference signals corresponding to a set of ink densities that can be output, and an outer cube 42 can take it. It shows conversion of luminance / color difference signals (luminance / color difference space). Therefore, outside the region of the curved body 41 of the curved surface in FIG. 4, the set of luminance / color difference signals does not correspond to the corresponding ink density, that is, the color that can be output, in the above calculation.

Therefore, according to the present invention, conversion to an ink density signal is performed even for a set of luminance / color difference signals which does not correspond to such an outputtable color.

This will be described with reference to FIGS. 4 and 5. FIG. 5 shows the space of the standardized color signals 6a, 6b, 6c. The area of the standardized color signal set (x1, x2, x3) corresponding to the set of ink densities that can be output by the curved surface body 51 in the figure is shown, and this area corresponds to the curved surface body 41 of FIG. Also, the outer cube 52 shows a region that the standardized color signal can take. FIG. 4 The conversion value of the point Y outside the area 41 of the curved surface body to the ink density value is determined as follows. That is, first, a set of standardized color signals corresponding to the luminance color difference signal Y (y1, y2,
y3) is obtained and the point is designated as Y '. And curved body 71
Error point E in the point of the inside, that is, in the set of normalized color signals (x1, x2, x3) corresponding to the set of printable ink densities

[0018]

[Equation 2] Then, a set (x1, x2, x3) that minimizes is obtained (Fig. 5 X '), and the converted value of the signal of the set of luminance color difference signals (X in Fig. 4) corresponding to this set of standardized color signals , And the converted value of the signal Y. Here, Vi is the variance of noise on the standardized color signal.

By performing this operation for all the signals outside the curved surface body 41 of FIG. 4, the contents of the conversion table 11 for converting the luminance color difference signals into the ink density signals can be filled without any gap.

The unique effects of constructing the conversion table 11 as described above are as follows. Generally, the color signal 2 contains noise, and it is considered that this noise is the cause of deviation from the region in which the luminance color difference signal or the standardized color signals 6a, 6b, 6c can be expressed.
It can be considered that the noises added to the color signal 2 have low correlation with each other, and the noises are approximately Gaussian.

Therefore, the noise added to the standardized color signal is also low in correlation with each other and is Gaussian. Therefore, when the set of luminance color difference signals deviates from the color range that can be output due to noise, that is, when the standardized color signal deviates from the color range that can output and becomes (y1, y2, y3),
The original set of normalized color signals without noise is (x1, x
The posterior probability P that is (2, x3) is

[0022]

[Equation 3] And the original signal is the one that maximizes P among the colors that can be expressed, that is,

[0023]

[Equation 4] It is reasonable to think of as minimizing.

With the configuration of the conversion table shown in this embodiment, when a luminance color difference signal that does not correspond to an outputable color is input, the signal with the highest likelihood is input as the original signal. Which is equivalent to, and a good color image can be reproduced. Further, although the noise variance Vi is used in the above embodiment, it may not always be possible to use such an expression due to the definition of the color signal space. Generally, a constant Vij that represents the statistical property of noise is used.

Of course, if there is no problem in the output image by the simplest signal correction, simpler correction is allowed. For example, one point in the space of the outputtable color signal may be arbitrarily selected, a straight line connecting this point and the input signal may be determined, and the intersection with the space of the outputtable color signal may be used as the correction value. At this time, how to select one point in the space may be appropriately determined depending on the read image.

[0026]

According to the present invention, even when noise is mixed in the image input section or a color signal that cannot be output is detected and a color signal that does not have a corresponding outputable color is input in principle. An image is output by obtaining an output signal for outputting a color that can be output close to the color corresponding to this color signal. Therefore, it is possible to obtain a high-quality color image having a good S / N visually and good color reproducibility.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of a color copying apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing a configuration of an image input unit in the embodiment.

FIG. 3 is a diagram showing input / output characteristics of a gamma correction circuit in the example.

FIG. 4 is a diagram showing an existing area of a set of luminance color difference signals corresponding to colors that can be output by a color printer in a luminance color difference space, and showing a configuration of a color conversion table.

FIG. 5 is a diagram showing an existing area of a set of standardized color signals corresponding to colors that can be output by a color printer in the standardized color signal space, and for explaining the configuration of a color conversion table.

[Explanation of symbols]

1 ... Image input section 2 ... Color signal 3 ... A / D conversion circuit 5 ... Shooting correction circuit 6 ... Normalized color signal 7 ... Luminance color difference separation circuit 8 ... Luminance signal 9a, 9b ... Color difference signal 10 ... Gamma correction circuit 11 ... Conversion table 12 ... Ink density signal 13 ... Color printer 21 ... Input image 22 ... Light source 23 ... Lens 24 ... Photodetector array 25 ... Color filter 41 ... Luminance / color difference signal group area corresponding to outputtable ink density group 42 ... Luminance color difference signal space 51 ... Standardized color signal set area corresponding to a set of printable ink densities 52 ... Normalized color signal space

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical indication H04N 9/64 Z 8626-5C

Claims (1)

[Claims] 1. A first step of inputting a plurality of detected color signals, and a step of converting the color signals input in the first step into output signals suitable for an output form of a color image output device. Two
And a step of obtaining a predetermined color signal corresponding to the color signal input in the first step, and the first step. If the color signal input in step 1 does not correspond to a color that can be output by the color image output device,
Among the predetermined color signals corresponding to the colors that can be output by the color image output device, a predetermined color signal that minimizes the error distance E from the predetermined color signal obtained in the third step is selected. A fourth step of obtaining, and a fifth step of making the output signal corresponding to the predetermined color signal obtained in the fourth step the output signal for the color signal input in the first step. A color image conversion method including: