JP4049763B2 - Color conversion table, color conversion device, color signal generation device, color conversion table creation method, color conversion method, and color gamut determination method - Google Patents

Description

  The present invention relates to a color conversion table, a color conversion device, a color signal generation device, a color conversion table creation method, a color conversion method, and a color gamut determination method, and in particular, accurately converts color signals expressed in different expression formats. It is suitable to be applied to.

  In recent years, full-colorization has rapidly progressed in printers, facsimiles, and the like, and there is a demand for sharing color images between different devices. However, when a color image is to be shared between different devices, the color reproducibility differs depending on the device, which causes a problem that the colors of the images obtained on the transmitting side and the receiving side are different.

  For example, printers and scanners handle colors in the form of color signals sent to devices such as CMY signals and RGB signals, respectively. That is, in a printer in which the minimum value of the output color signal is 0 and the maximum value is 255, all output colors are expressed by combinations of values in the range of 0 to 255 for each CMY signal. In this case, even if CMY signals having the same value are input, the color to be printed differs depending on the printer model. Therefore, the color expression method based on the CMY signals can be said to be a color representation method dependent on the device. .

  Therefore, conventionally, the color reproducibility between different devices has been made consistent by representing the color signal represented in a device-dependent representation format in an intermediate representation format independent of the device. Yes.

For example, as a method of expressing a color representing an absolute color that does not depend on a device, there is a method of expressing a color using an L ^* a ^* b ^* signal or an XYZ signal. Since the colors represented in these formats are uniquely defined, they are used as intermediate color representation methods when color conversion is performed between the color spaces of different devices.

Specifically, the image expressed by RGB signals inputted from the scanner, converted into L ^* a ^* b ^* signal image represented by, further, converts the L ^* a ^* b ^* signal to the CMY signal Then output to the printer. If conversion of the expression format of these color signals is accurately performed, colors can be matched between devices.

  In order to convert the representation format of the color signal, it is necessary to convert the color signal value represented in each format into the color signal value of another format. As a method of converting the value of the color signal, there is a method of using a color conversion table showing the correspondence of colors in each color space. In the color signal conversion method using the color conversion table, when converting the color signal registered in the color conversion table, conversion is performed using the correspondence registered in the color conversion table. When converting a color signal not registered in the color conversion table, the color signal to be converted is converted by interpolating the color signal registered in the color conversion table.

FIG. 38 is a diagram for explaining a conventional color conversion table registration method two-dimensionally on a color space.
In Figure ^{38, L * a * b *} L * a * b * values of grid points P1~P16 space corresponds to the CMY value of the point Q1~Q16 the CMY space, the color conversion table, L ^* a ^* b ^* It is assumed that the correspondence relation about the grid point is registered. Here, assuming that the CMY values of the points Q1, Q2, Q5, Q9, and Q13 exceed the color reproduction range of the device, in the conventional color conversion table, the point Q1, which exceeds the color reproduction range of the device, The CMY values of Q2, Q5, Q9, and Q13 are registered by being replaced with CMY values of points Q1 ′, Q2 ′, Q5 ′, Q9 ′, and Q13 ′ within the device color reproduction range.

Hereinafter, a color conversion method using a conventional color conversion table will be described more specifically using numerical examples. Here, printer color conversion (conversion from L ^* a ^* b ^* signal to CMY signal) using a value in the range from 0 to 255 (CMY value) as a color reproduction range is taken as an example. Note that the color signal may be a device other than a printer, or a color signal other than an L ^* a ^* b ^* signal or a CMY signal.

FIG. 39 is a table showing a configuration of a conventional color conversion table. This color conversion table shows the correspondence between the L ^* a ^* b ^* color signal and the CMY color signal. By using this color conversion table, the L ^* a ^* b ^* color signal is converted into the color gamut of the printer. Conversion to CMY color signals is possible.

In FIG. 39, the correspondence relationship between the L ^* a ^* b ^* color signal and the CMY color signal in the printer color gamut is 0 ≦ L ^* ≦ 100, −128 ≦ a ^* ≦ 128, −128 ≦ b ^* ≦ 128. In the range, data is discretely registered at intervals of 12.5 for the L ^* signal and 32 intervals for the a ^* signal and the b ^* signal. This is because it is practically impossible to register in the color conversion table so as to cover all values. Also, in the color conversion table of FIG. 39, the distribution of L ^* a ^* b ^* values is a regular value (an equally spaced value).

Therefore, the CMY values corresponding to the L ^* a ^* b ^* values registered in the color conversion table can be directly obtained from the color conversion table.
On the other hand, the CMY values corresponding to the L ^* a ^* b ^* values not registered in the color conversion table are interpolated using the correspondence relationship between the L ^* a ^* b ^* values and the CMY values registered in the color conversion table. Calculate by calculation. In this case, if the distribution of the L ^* a ^* b ^* values registered in the color conversion table is made regular, the interpolation calculation becomes easy.

For example, when L ^* a ^* b ^* value = (0, −128, −64) is input, L ^* a ^* b ^* value = (0, −128, −64) is the color conversion table of FIG. Therefore, CMY values corresponding to L ^* a ^* b ^* values = (0, −128, −64) = (255, 240, 253)) are referred to the color conversion table of FIG. Can be obtained directly.

On the other hand, for example, when L ^* a ^* b ^* value = (1, −128, −64) is input, L ^* a ^* b ^* value = (1, −128, −64) is the color shown in FIG. Since it is not registered in the conversion table, CMY values corresponding to L ^* a ^* b ^* values = (1, −128, −64) are obtained by interpolation calculation.

Here, in the conventional color conversion table, the CMY values registered in the color conversion table are only values within the range of the color signal of the target device. That is, when the color conversion table is used for a printer that targets CMY values in the range of 0 to 255, CMY values corresponding to L ^* a ^* b ^* values outside the printer's color gamut are set to 0 to 255. The color conversion table is registered after the CMY value is replaced.

For example, it is assumed that black that can be reproduced by the target printer has an L ^* value in the range of up to about 18 and cannot output darker colors. In this case, for the L ^* a ^* b ^* value in the color gamut of the printer, a CMY value having the same color reproducibility as the color indicated by the L ^* a ^* b ^* value is registered. For example, L ^* a ^{For *} b ^* value = (25, 0, 0), 228, 224, 208) are registered as CMY values.

Also, for L ^* a ^* b ^* values outside the printer's color gamut, the color indicated by the L ^* a ^* b ^* values is registered with CMY values that can be reproduced by the printer. For example, as a CMY value corresponding to L ^* a ^* b ^* value = (12.5, 0, 0), black of L ^* = 12.5 cannot be output. (As dark as possible) is selected and (255, 255, 255) is registered.

When converting L ^* a ^* b ^* values not registered in the color conversion table into CMY values, the correspondence between the L ^* a ^* b ^* values and the CMY values stored in the color conversion table is as follows: A plurality of correspondences around the L ^* a ^* b ^* values to be converted are selected, and the selected correspondences are obtained by interpolation. The interpolation method is described in Non-Patent Document 1, for example.

Hereinafter, a color signal interpolation method using a conventional color conversion table will be described using numerical examples.
In color conversion by interpolation using a conventional color conversion table, if the values registered in the color conversion table are accurate, generally good color conversion results can be obtained by color conversion by interpolation within the device gamut. Is possible.

Here, L ^* a ^* b ^* value = (25,0,0) and L ^* a ^* b ^* value = (37.5,0,0) registered in the color conversion table of FIG. A case where CMY values corresponding to L ^* a ^* b ^* values = (31.25, 0, 0) that are not registered in the color conversion table of FIG. (Note that in practice, interpolation is performed three-dimensionally using a larger number of colors.)

In the interpolation operation, basically, an operation for obtaining a color having the same positional relationship between the L ^* a ^* b ^* signal and the CMY signal is performed. Therefore, L ^* a ^* b ^* value = (25, 0, 0) and L ^* a ^* b ^* value = the two data of (37.5,0,0) intermediate L ^* a ^* b ^* value The CMY values corresponding to L ^* a ^* b ^* values = (25, 0, 0) are calculated as CMY values = (228, 224, 208) and L ^* a ^* b ^* values = (37 .5,0,0), CMY value = (182,185,161) is obtained, and CMY value intermediate between CMY value = (228,224,208) and CMY value = (182,185,161) Is a process for obtaining.

Therefore, the CMY value of the interpolation calculation result is (205, 204.5, 184.5). That is, the conversion result to the CMY color signal of L ^* a ^* b ^* value = (31.25, 0, 0) not registered in the color conversion table is CMY value = (205, 204.5, 184.5). ).

  The result of this interpolation calculation is generally correct if there are no extreme irregularities in the printer characteristics. This is because if the correspondence relationship between the colors registered in the color conversion table is accurate, the values positioned between the correspondence relationships will be substantially accurate.

In the conventional color conversion table, colors that cannot be reproduced by the device are registered by replacing them with the closest colors that can be reproduced by the device ^. It was not possible to determine whether the a ^* b ^* value could be reproduced by the device or not. For this reason, an identifier indicating whether the color gamut of the device is inside or outside the color gamut is registered for each L ^* a ^* b ^* value of the grid-like distribution (The ICC Profile Specifications Version 3.3), or at a certain interval L. In some cases, the L ^* a ^* b ^* value is determined to be within the device gamut or out of the gamut by registering the locus of the gamut and using it as a boundary descriptor.
"High-precision color conversion by flexible UCR" (JAPAN Hard Copy paper), 1994, p. 177-180

  However, when the conventional color conversion table is used, accurate interpolation calculation (color conversion processing) is difficult for colors that are close to the color gamut of the device (bright colors, very dark colors, very light colors). There was a problem.

For example, in FIG. 38, when the CMY value corresponding to the point R near the color gamut boundary in the L ^* a ^* b ^* space is obtained, the CMY value corresponding to the point R is not registered in the color conversion table. Perform interpolation calculation. In this interpolation calculation, CMY values corresponding to the grid points P2, P3, P6, and P7 surrounding the point R are obtained from the color conversion table, and a weight calculation is performed on these CMY values to obtain the CMY value corresponding to the point R. Ask. In this case, although the CMY value corresponding to the lattice point P2 is actually the CMY value of the point Q, it is registered in the color conversion table in place of the CMY value of the point Q ′. Therefore, although the actual CMY value corresponding to the point R is the CMY value of the point S, the CMY value of the point S ′ is calculated, and as a result, a CMY value different from the actual CMY value is output. The

Hereinafter, the case where the L ^* a ^* b ^* value to be interpolated is in the vicinity of the color gamut boundary will be described using specific numerical values registered in the color conversion table.
Assume that the target printer can output up to L ^* a ^* b ^* value = (18, 0, 0) for dark colors. In this case, as shown in FIG. 39, since the color corresponding to L ^* a ^* b ^* value = (12.5, 0, 0) cannot be output by this printer, it can be output by this printer as a CMY value. The CMY values (255, 255, 255) of the darkest color are registered.

Here, L ^* a ^* b ^* value = (18.75, 0, 0) (L ^* a ^* b ^* value = (12.5, 0, 0) and L ^* a ^* b ^* value = (25, It is assumed that a CMY color signal corresponding to the middle of 0,0) is obtained by interpolation.

As already described, since the target printer can output up to L ^* a ^* b ^* value = (18, 0, 0), by specifying CMY value = (252, 253, 235), L ^* A color of a ^* b ^* value = (18.75, 0, 0) can be output. When printing a color of L ^* a ^* b ^* value = (18.75, 0, 0) with a printer capable of printing up to a color of L ^* a ^* b ^* value = (18, 0, 0), here Like the set condition, the CMY value is a value close to 255.

Therefore, it is desirable to calculate CMY value = (252, 253, 235) as an interpolation calculation result for L ^* a ^* b ^* value = (18.75, 0, 0).
However, using the color conversion table of FIG. 39, CMY values corresponding to L ^* a ^* b ^* values = (12.5, 0, 0) = (255, 255, 255) and L ^* a ^* b ^* = By averaging (228,224,208) corresponding to (25,0,0), the CMY value corresponding to the intermediate L ^* a ^* b ^* value = (18.75,0,0) , CMY value = ((228 + 255) / 2, 224 + 255) / 2, 208 + 255) / 2) = (241.5, 239.5, 231.5), and there is a large deviation from the actual value. Become.

This is because, in the conventional color conversion table, for L ^* a ^* b ^* value = (12.5, 0, 0), although it corresponds to a CMY value outside the device color gamut, The CMY values at the region boundaries are registered. For this reason, when the interpolation calculation is performed using these two colors, the color near the color gamut boundary is actually calculated as a color substantially inside the color gamut.

  In addition, regarding the conventional method of determining whether the color gamut is outside or outside the color gamut of the device, the method using the identifier registered in the grid points cannot determine the color gamut with a resolution finer than the grid point interval. It was. In the method of registering the locus for each hue, in order to use the locus that is divided and registered for each hue, an operation (trigonometric function) for obtaining the hue is necessary and complicated.

Accordingly, a first object of the present invention is to make it possible to accurately perform color conversion near the color gamut boundary.
A second object of the present invention is to easily determine whether or not a color signal to be converted can be reproduced.

  In order to solve the above-described problem, according to the present invention, the correspondence relationship between the color signal of the first color space and the color signal of the second color space is stored in the color conversion table beyond the color reproducible range. I am trying to register.

  As a result, even when the color indicated by the color signal in the first color space exceeds the color reproducible range, the color indicated by the color signal in the first color space is stored and the second color space is preserved. The color signal can be registered in the color conversion table. For this reason, it is possible to prevent the loss of accuracy even when the correspondence relationship of the color signal exceeding the color reproducible range is used when obtaining the correspondence relationship of the color signal not registered in the color conversion table. It becomes possible.

According to another aspect of the invention, the first color space is an independent color space that does not depend on a device, and the second color space is a device-dependent color space that depends on a device.
As a result, even when color signals are transmitted and received between devices having different color reproducibility, the color reproducibility between the devices can be matched.

According to one aspect of the present invention, the distribution of the color signal values in the first color space is regular.
This makes it possible to easily select the color signal used in the interpolation process and perform the interpolation calculation when obtaining the correspondence relationship of the color signal not registered in the conversion table by the interpolation process, and to perform the color conversion at high speed. Can be done.

Also, according to one aspect of the present invention, an identifier for identifying a color signal that is a predetermined value or more away from a color reproducible range is registered in the color conversion table.
As a result, when a color signal outside the color reproducible range is generated by a method such as extrapolation based on actual measurement values, the influence of errors increases as the extrapolation distance increases. It is possible to easily identify a color signal that is significantly out of accuracy with respect to a color signal that deviates greatly.

In addition, according to one aspect of the present invention, an identifier for identifying whether a color signal is within or outside a color reproducible range is registered in the color conversion table.
This makes it possible to determine the color gamut of the color signal simply by referring to the color conversion table when determining the color gamut of the color signal, and to perform the color gamut determination of the color signal at high speed. It becomes.

Further, according to one aspect of the present invention, an identifier for identifying whether a color signal is near the color reproduction boundary is registered in the color conversion table.
This makes it possible to determine the color gamut of the color signal by actually performing color conversion only when the identifier is included when determining the color gamut of the color signal. Color gamut determination can be performed at high speed.

Further, according to one aspect of the present invention, it is provided with the interpolating unit that interpolates the color signal based on the color signal of the second color space incapable of color reproduction.
Thus, it is possible to perform color signal interpolation without being affected by the color reproduction range of the device, and it is possible to prevent the accuracy in interpolating the color signal from being lost.

Further, according to the aspect of the invention, when the interpolation result is not within the color reproducible range, the replacement unit is provided that replaces the interpolation result with the color signal of the second color space capable of color reproduction.
As a result, it is possible to prevent a color signal that cannot be handled by the device from being output to the device, and that the color signal cannot be processed by the device.

  Further, according to one aspect of the present invention, the correspondence relationship between the color signal of the grid point on the first color space and the color signal on the second color space is registered beyond the color reproduction range of the device. By referring to the color conversion table, the correspondence relationship of the color signals not registered in the color conversion table is obtained by interpolation.

  Thus, even when a color signal of the first color space not registered in the color conversion table is input and the color signal of the first color space is near the boundary of the color reproduction range, the color reproduction of the device is performed. By using a color signal that exceeds the range, it is possible to interpolate the color signal near the boundary of the color reproduction range, and it is possible to prevent the accuracy when interpolating the color signal from being lost.

  In addition, according to one aspect of the present invention, a color signal that is out of the color reproduction range is moved by an excess level from the color reproduction range in the direction of an achromatic signal having the same luminance as the color signal. I have to.

  As a result, when the color signal is obtained by interpolation, even if the color signal falls outside the color reproduction range, the color signal within the color reproduction range that is closest to the color signal obtained by interpolation. Therefore, the color reproducibility of the color signal obtained by the interpolation and the color signal obtained by the replacement can be substantially matched. In addition, the color signal between the color signal that is out of the color reproduction range and the color signal of the same luminance and achromatic color is the color signal at the boundary of the color reproduction range. It is possible to calculate by this simple method.

  According to one embodiment of the present invention, as a color signal of a device in a color gamut, when the luminance of the color signal to be converted is high, a device corresponding to a color equal to or lower than the luminance of the color signal to be converted When the luminance of the color signal to be converted is low, the color signal of the device corresponding to the color equal to or higher than the luminance of the color signal to be converted is selected.

  As a result, since the color reproduction area has a wide middle luminance, it is possible to replace the color signal to be converted with a color close to the color signal by a simple method such as linear interpolation.

Further, according to one aspect of the present invention, when the color signal used for interpolation is not registered in the color conversion table, the saturation of the color signal to be converted is reduced.
As a result, even when the color signal to be converted cannot be interpolated, the color signal that cannot be interpolated is converted into a color signal that can be interpolated while preserving the luminance and hue of the color signal to be converted. The color conversion can be performed while maintaining the color reproducibility to some extent.

  According to one aspect of the present invention, even when the saturation of a color signal to be converted is set to 0, if the color signal used for interpolation is not registered in the color conversion table, the color to be converted The signal brightness is changed to the inside.

As a result, it is possible to detect the color signal used for interpolation using the color conversion table in a wide color reproduction range, and to find the color signal used for interpolation.
According to another aspect of the present invention, color signal selection means for selecting a color signal capable of color reproduction and color signal extrapolation means for extrapolating the selected color signal are provided.

This makes it possible to calculate a color signal for a color that cannot be actually measured, and to generate a color signal outside the color reproduction range of the device.
Further, according to one aspect of the present invention, the correspondence relationship outside the color reproduction range of the device is obtained based on the correspondence relationship of the color signals obtained within the color reproduction range of the device.

  This makes it possible to calculate the correspondence of the color signal even for colors that are not actually reproducible, and to register the correspondence outside the device color reproduction range in the color conversion table.

  Further, according to one aspect of the present invention, a first point adjacent to the outside of the boundary of the device color reproduction range is selected on the first color space, and the first point is selected on the first color space. Select a second point adjacent to the inside of the point, select a third point that is one inward from the second point, and a fourth point on the second color space corresponding to the second point And a fifth point on the second color space corresponding to the third point, and a sixth point that is point-symmetric with the fifth point about the fourth point on the second color space. The color signal of the sixth point is obtained, and the color signal of the sixth point is registered in the color conversion table in correspondence with the color signal of the first point.

  This makes it possible to create a color signal outside the color gamut using a color signal in the color gamut located near the color signal to be created, and to create a color signal with high accuracy outside the color gamut. Become. In addition, since the color signal created outside the color gamut can be calculated by simple calculation using the color signals of two points within the color gamut, the color signal outside the color gamut can be calculated at high speed. Become.

Moreover, according to one aspect of the present invention, the extrapolated position is changed based on the distance between two points used for extrapolation.
This makes it possible to specify the position of the color signal set outside the color gamut, and widen or narrow the interval of the color signal set outside the color gamut and the setting range of the color signal set outside the color gamut. Therefore, it is possible to control the distribution state of the color signals set outside the color gamut, and to efficiently set the color signals outside the color gamut.

According to one aspect of the present invention, the points created by extrapolation are generated on a straight line connecting two points used for extrapolation.
This makes it possible to create a correspondence relationship outside the color gamut from the correspondence relationship within the color gamut by linearly extrapolating the color signals within the color gamut, and easily create a correspondence relationship outside the color gamut. It becomes.

  Further, according to one aspect of the present invention, by evaluating both or either of the distance between two points used in extrapolation and the direction of a straight line connecting the two points used in extrapolation, The point used in is changed to a more internal point in the color gamut.

  Because of this, color collapse occurs in the color around the gamut boundary, so even if you change the value of the color signal near the gamut boundary, the color reproduced by the device will not change much, etc. Even if it is inappropriate to use extrapolation because there is almost no difference in color between the two points used in extrapolation, change the point used in extrapolation to a point with a larger color difference. Therefore, it is possible to maintain accuracy when generating a color signal out of the color gamut.

  Further, according to one aspect of the present invention, the first correspondence between the color signal of the first color space and the color signal of the second color space is obtained within the color reproduction range of the device, and the first correspondence is obtained. Based on the relationship, a second correspondence relationship between the color signal of the first color space and the color signal of the second color space is obtained outside the color reproduction range of the device, and the first correspondence relationship and the second correspondence relationship are obtained. Based on the relationship, a third correspondence relationship between the color signal of the first color space and the color signal of the second color space is obtained, and the third correspondence relationship is registered in the color conversion table.

  This makes it possible to generate a new color signal correspondence from the correspondence between the color signals in the color gamut and outside the generated color gamut, increase the number of color signals registered in the color conversion table, It is possible to replace it with a signal, to generate a correspondence relationship between color signals outside the generated color gamut, and to change the distribution of color signals within the color gamut and outside the generated color gamut.

  Further, according to one aspect of the present invention, the first correspondence relationship between the color signal irregularly distributed in the first color space and the color signal distributed at the grid points of the second color space is obtained. By obtaining the device within the color reproduction range of the device and interpolating the first correspondence relationship, the color signal distributed at the lattice points of the first color space and the second color space are irregularly distributed. A second correspondence relationship with a color signal that is present is determined within the color reproduction range of the device, and the second correspondence relationship is extrapolated to obtain a color signal distributed in the grid of the first color space and the first correspondence relationship. The third correspondence relationship with the color signal irregularly distributed in the color space 2 is obtained outside the device color reproduction range, and the second correspondence relationship and the third correspondence relationship are registered in the color conversion table. Like to do.

  As a result, the color signal distributed in the first color space can be directly registered in the color conversion table, and the color signal in the first color space is distributed at the lattice points. By sequentially selecting and extrapolating the color signals of the first color space, extrapolation points of the color signals of the first color space are generated at uniform intervals, and the efficiency exceeds the color reproduction range of the device. It is possible to generate a color conversion table capable of performing color conversion at high speed.

  Further, according to one aspect of the present invention, the first correspondence relationship between the color signal irregularly distributed in the first color space and the color signal distributed at the grid points of the second color space is obtained. By obtaining the device within the color reproduction range of the device and extrapolating the first correspondence relationship, the color signal is irregularly distributed in the first color space and distributed to the grid points of the second color space. The second correspondence relationship with the color signal is obtained outside the color reproduction range of the device, and the first correspondence relationship and the second correspondence relationship are interpolated to be distributed to the grid points of the first color space. A third correspondence relationship between the color signal distributed irregularly in the second color space and the color signal irregularly distributed in the second color space is determined within the device color reproduction range and outside the device color reproduction range. Are registered in the color conversion table.

  As a result, the first correspondence can be directly extrapolated to obtain the second correspondence outside the color reproduction range of the device, and the first correspondence can be accurately obtained by actual measurement. From this, it is possible to prevent the accumulation of errors when obtaining the second correspondence relationship outside the color reproduction range of the device, and by performing interpolation after obtaining the second correspondence relationship. The color signal distributed in the first color space can be distributed to the grid points, and a color conversion table capable of efficiently performing color conversion beyond the device color reproduction range is generated with high accuracy. It becomes possible.

  According to one embodiment of the present invention, a unit cube surrounded by lattice points in the second color space is divided into a plurality of cubes, and the color signal in the second color space of the vertexes of the divided cubes is represented as a vertex. The color signal on the first color space corresponding to the vertex of the divided cube is calculated by the interpolation calculation process, and the value of the color signal calculated by the interpolation calculation process is the first color space. When the value is within a predetermined range from the value of the grid point on the color space, the color signal on the second color space at the vertex of the divided cube corresponding to the color signal of the grid point on the first color space To choose.

  As a result, the color signal distributed irregularly in the first color space and the color signal in the second color space are distributed to the lattice points in the first color space. It is possible to approximately obtain the correspondence between the color signal and the color signal in the second color space, and by increasing the number of divisions of the unit cube, the value of the color signal calculated by the interpolation calculation process can be obtained. Since it is possible to approximate the value of the grid point on the first color space, the correspondence between the color signal distributed at the grid point on the first color space and the color signal on the second color space It becomes possible to improve the accuracy of the relationship.

  Further, according to one aspect of the present invention, the volume of a rectangular parallelepiped formed when the unit cube is divided at the vertex of the divided cube is calculated on the second color space, and is calculated on the second color space. The volume of the rectangular parallelepiped is used as a weighting coefficient when a color signal in the first color space is obtained by interpolation calculation processing.

  As a result, since the color signals are regularly distributed in the second color space, the volume of the rectangular parallelepiped can be easily obtained, and the volume of the rectangular parallelepiped obtained in the second color space can be calculated. By using it as a weighting coefficient in one color space, a color signal in the first color space can be efficiently obtained by interpolation calculation processing.

  According to one embodiment of the present invention, a first color signal is input, a second color signal in the vicinity of the first color signal is selected, and the color signal can be handled beyond the color reproduction range of the device. The correspondence relationship between the second color signal and the third color signal is detected by referring to the color conversion table in which the relationship is registered, and the correspondence relationship between the second color signal and the third color signal is detected. Based on this, the first color signal is converted into a fourth color signal.

  As a result, when the first color signal is converted into the fourth color signal, the first color signal is not registered in the color conversion table, so the second color signal in the vicinity of the first color signal. When the color conversion is performed by selecting the color conversion table from the color conversion table, the correspondence relationship for the second color signal is not restricted by the color reproduction range of the device. It is possible to perform conversion to.

  According to one aspect of the present invention, a second color signal in the vicinity of the first color signal is selected, and the second color signal includes a value outside the color reproduction range of the device. To determine whether the first color signal is included in the device color reproduction range by checking whether the third color signal is included in the device color reproduction range. I have to.

  As a result, the first color signal not registered in the color conversion table is input, and the color reproduction range of the device is irregular in the color space to which the first color signal belongs. When it is difficult to obtain the range, it is possible to obtain the color gamut of the second color signal registered in the color conversion table with the boundary of the device color reproduction range being regular and the device color reproduction range. Since it is possible to easily evaluate the color space of the third color signal and to determine the color gamut of the second color signal as the color gamut of the first color signal, It is possible to efficiently perform the color gamut determination of the first color signal that is not registered in the color conversion table.

  According to one aspect of the present invention, a second color signal in the vicinity of the first color signal is selected, and the second color signal includes a value outside the color reproduction range of the device. And converting the first color signal into a fourth color signal based on the third color signal, and determining whether the fourth color signal is within the color reproduction range of the device. Thus, it is determined whether or not the first color signal is within the color reproduction range of the device.

  As a result, the first color signal not registered in the color conversion table is input, and the color reproduction range of the device is irregular in the color space to which the first color signal belongs. When it is difficult to obtain the range, the first color signal can be interpolated based on the second color signal registered in the color conversion table, and the boundary of the color reproduction range of the device Since the color gamut can be determined in a color space that is regular and it is easy to obtain the device color reproduction range, the color gamut determination of the first color signal not registered in the color conversion table is possible. Can be performed efficiently.

  According to one aspect of the present invention, when the converted third color signal includes only a value within the color reproduction range of the device, the first color signal before conversion is the color reproduction range of the device. If the converted third color signal includes only a value outside the color reproduction range of the device, the first color signal before conversion is outside the color reproduction range of the device. Judgment is made.

  As a result, even when a first color signal not registered in the color conversion table is input, the first color signal is within the color reproduction range of the device without interpolation of the first color signal. Therefore, it is possible to determine whether it is out of the color reproduction range of the device, and it is possible to perform color gamut determination at high speed.

  As will be described below, according to the present invention, the correspondence relationship of color signals not registered in the color conversion table is interpolated by registering the correspondence relationship of color signals beyond the color reproducible range in the color conversion table. Thus, even when a color signal correspondence relationship exceeding the color reproducible range is used, it is possible to prevent the accuracy during interpolation from being impaired.

  Further, according to one aspect of the present invention, the correspondence relationship between the device-independent independent color space and the device-dependent color space is registered in the color conversion table beyond the color reproducible range of the device. As a result, even when color signals are transmitted and received between devices having different color reproducibility, the color reproducibility between the devices can be matched.

  Further, according to one aspect of the present invention, by making the distribution of color signal values in an independent color space independent of the device regular, the correspondence relationship of color signals not registered in the conversion table can be interpolated. When obtaining, it is possible to easily select a color signal to be used in the interpolation process and to perform an interpolation calculation, and to perform color conversion at high speed.

  In addition, according to one aspect of the present invention, by registering an identifier of a color signal that is a predetermined value or more away from the color reproducible range in the color conversion table, the color signal outside the color reproducible range can be measured. If the extrapolation distance increases, the effect of the error increases, and it is easy to identify that the color signal that is greatly out of the color gamut may not be accurate. It becomes possible to do.

  Further, according to one aspect of the present invention, an identifier for identifying whether a color signal is within a color reproducible range or out of the range is registered in the color conversion table, so that when determining the color gamut of the color signal, It is possible to determine the color gamut of the color signal simply by referring to the conversion table, and it is possible to determine the color gamut of the color signal at high speed.

  According to one aspect of the present invention, an identifier for identifying whether or not a color signal is near the color reproduction boundary is registered in the color conversion table, so that when color gamut of the color signal is determined, color conversion is actually performed. It is possible to easily determine whether or not it is necessary to perform the determination, and it is possible to efficiently perform the color gamut determination of the color signal.

  Further, according to one aspect of the present invention, by performing color signal interpolation based on a color signal that cannot be reproduced, the color signal can be interpolated without being affected by the color reproduction range of the device. Thus, it is possible to prevent the accuracy when interpolating the color signal from being lost.

  Further, according to one aspect of the present invention, when the interpolation result is not within the color reproducible range, the color signal that cannot be handled by the device is replaced by replacing the interpolation result with a color signal that can be reproduced. It is possible to prevent the color signal from being processed by the device.

  According to one embodiment of the present invention, a color signal that is out of the color reproduction range is a color signal at the boundary of the color reproduction range among color signals between the color signal and an achromatic color signal having the same luminance. Can be replaced by a simple method such as linear interpolation.

  Further, according to one aspect of the present invention, it is possible to calculate a color signal for a color that cannot be measured by extrapolating a color signal that can be reproduced, so that the color signal is outside the color reproduction range of the device. Can be generated.

  In addition, according to one embodiment of the present invention, it is possible to create a color signal with high accuracy outside the color gamut by extrapolating using the color signal in the color gamut adjacent to the color signal to be created. Since the color signal created outside the color gamut can be calculated by a simple calculation using the color signals of two points within the color gamut, the color signal outside the color gamut can be calculated at high speed. .

  Further, according to one aspect of the present invention, the distribution state of the color signal set outside the color gamut by changing the extrapolated position based on the distance between two points used for extrapolation. Therefore, it is possible to efficiently set color signals out of the color gamut.

According to one embodiment of the present invention, a color signal outside the color gamut can be easily generated by linearly extrapolating the color signal within the color gamut.
In addition, according to one aspect of the present invention, when there is almost no difference in color between two points used in extrapolation, when it is determined that the error of a point created by extrapolation is large, extrapolation is performed. By changing the point to be used to a point having a larger color difference, it is possible to accurately create a color signal outside the color gamut.

  Further, according to one aspect of the present invention, by generating a new color signal correspondence from the correspondence between the color signals in the color gamut and outside the generated color gamut, the number of color signals to be registered in the color conversion table can be increased. It is possible to replace the color signal with a convenient color signal, generate a correspondence relationship between the color signals outside the generated color gamut, and change the distribution of the color signals within the color gamut and outside the generated color gamut. .

  In addition, according to one aspect of the present invention, the device-independent color space is generated by extrapolating the device-independent color space grid points to generate a color signal correspondence outside the device color reproduction range. Extrapolation points can be generated at uniform intervals, and a color conversion table capable of efficiently performing color conversion beyond the color reproduction range of the device can be generated at high speed.

  In addition, according to one aspect of the present invention, after obtaining the correspondence relationship of the color signals outside the device color reproduction range, the interpolation processing is performed so that the color signals in the device-independent color space are distributed at the grid points. Accordingly, it is possible to accurately generate a color conversion table capable of efficiently performing color conversion beyond the device color reproduction range.

  Further, according to one aspect of the present invention, the correspondence relationship of the color signals is increased by interpolation on the regularly distributed second color space, and from these, the one close to the lattice point on the first color space Is obtained from the corresponding relationship between the color signal irregularly distributed in the first color space and the color signal in the second color space, and distributed to the grid points in the first color space. It is possible to approximately obtain the correspondence relationship between the existing color signal and the color signal in the second color space, and by increasing the number of interpolation points, the correspondence is distributed to the lattice points in the first color space. It is possible to improve the accuracy of the correspondence between the color signal and the color signal in the second color space.

  According to one aspect of the present invention, the weighting factor calculated in the second color space that is regularly distributed is used to obtain a color signal in the first color space that is irregularly distributed. Therefore, even when it is difficult to calculate the weighting coefficient in the first color space, the color signal in the first color space can be efficiently obtained by the interpolation calculation process.

  Further, according to one aspect of the present invention, the first color signal is changed to the fourth color signal by referring to the color conversion table in which the correspondence relationship of the color signal is registered beyond the color reproduction range of the device. At the time of conversion, since the first color signal is not registered in the color conversion table, the second color signal in the vicinity of the first color signal is selected from the color conversion table and color conversion is performed. Since the correspondence relationship with respect to the second color signal can be eliminated from being restricted by the color reproduction range of the device, the conversion from the first color signal to the fourth color signal can be performed with high accuracy. It becomes possible.

  According to one aspect of the present invention, when a first color signal that is not registered in the color conversion table is input, the color gamut of the second color signal that is registered in the color conversion table is determined by the device. The boundary of the color reproduction range is regular, and the color reproduction range of the device is evaluated on the color space of the third color signal, and the color gamut of the second color signal is the first color signal. Even if it is difficult to obtain the device color reproduction range in the color space to which the first color signal belongs by determining the color gamut, the boundary of the device color reproduction range is irregular. It becomes possible to efficiently perform the color gamut determination of the first color signal not registered in the table.

  According to one embodiment of the present invention, when a first color signal that is not registered in the color conversion table is input, the first color is based on the second color signal that is registered in the color conversion table. Interpolate the signal, and for the fourth color signal obtained as a result of this interpolation, the boundary of the color reproduction range of the device is regular, and the color gamut determination is performed in a color space where the device color reproduction range can be easily obtained. As a result, in the color space to which the first color signal belongs, even when it is difficult to obtain the device color reproduction range because the boundary of the device color reproduction range is irregular, it is registered in the color conversion table. It is possible to efficiently perform the color gamut determination of the first color signal that has not been performed.

Embodiments of the present invention will be described below with reference to the drawings.
FIG. 1 is a block diagram showing the configuration of a color conversion table according to an embodiment of the present invention.
In FIG. 1, the color conversion table 1 includes a first color conversion result 2 for converting an input color signal into a color signal that can be reproduced, and an input color signal that is converted into a color signal that cannot be reproduced. The second color conversion result 3 to be stored is stored. By registering color signals that cannot be reproduced in the color conversion table 1, the correspondence between the color signals can be accurately maintained even in areas beyond the color reproducible range such as printers and displays. When obtaining the correspondence of color signals not registered in Table 1, even when the correspondence of color signals exceeding the color reproducible range is used in the interpolation computation, the interpolation computation can be performed with high accuracy. .

FIG. 2 is a diagram showing a correspondence relationship between the L ^* a ^* b ^* signal and the CMY signal in a two-dimensional color space according to an embodiment of the present invention.
In FIG. ^{2, L * a * b *} L * a * b * values of grid points P1~P16 space corresponds to the CMY value of the point Q1~Q16 the CMY space, the color conversion table, L ^* a ^* b ^* It is assumed that the correspondence relation about the grid point is registered. Here, as points corresponding to L ^* a ^* b ^* values of lattice points P1, P2, P5, P9, and P13 in the L ^* a ^* b ^* space, points Q1, Q2, which exceed the device color reproduction range, The CMY values of Q5, Q9, and Q13 are registered as they are in the color conversion table.

Here, for example, when obtaining the CMY value corresponding to the point R near the color gamut boundary in the L ^* a ^* b ^* space, the CMY value corresponding to the point R is not registered in the color conversion table. CMY values corresponding to grid points P2, P3, P6, and P7 surrounding R are obtained from the color conversion table, and weight calculation is performed on these CMY values to obtain the CMY value of point Q corresponding to point R. To do.

In this case, as the CMY value corresponding to the grid point P2, the CMY value of the point Q2 exceeding the color reproduction range of the device is registered as it is in the color conversion table, so that L ^* a ^* b ^{* of the} grid point P2 Interpolation can be performed using the original color signal corresponding to the value, and the CMY value of the point S can be accurately calculated as the CMY value corresponding to the point R.

FIG. 3 is a table showing numerical examples registered in the color conversion table according to the embodiment of the present invention.
In FIG. 3, the correspondence relationship between the L ^* a ^* b ^* color signal and the CMY color signal in the printer color gamut is 0 ≦ L ^* ≦ 100, −128 ≦ a ^* ≦ 128, −128 ≦ b ^* ≦ 128. In the range, data is discretely registered at intervals of 12.5 for the L ^* signal and 32 intervals for the a ^* signal and the b ^* signal.

  Here, assuming that the range of CMY values in the printer gamut is from 0 to 255, in this color conversion table, CMY values are registered beyond the range of CMY values in the printer gamut.

For example, it is assumed that black that can be reproduced by the target printer has an L ^* value in the range of up to about 18 and cannot output darker colors. In this case, for the L ^* a ^* b ^* value in the color gamut of the printer, a CMY value that matches the color reproducibility with the color indicated by the L ^* a ^* b ^* value is registered. For example, L ^* a ^* b ^{For *} value = (25, 0, 0), (228, 224, 208) is registered as the CMY value.

Also, as a CMY value corresponding to L ^* a ^* b ^* value = (12.5, 0, 0) outside the printer color gamut, this printer cannot output black with L ^* = 12.5. The CMY values (273, 271, 255) corresponding to the L ^* a ^* b ^* values = (12.5, 0, 0) are registered in the color conversion table as they are.

By using this color conversion table, for example, CMY values corresponding to L ^* a ^* b ^* values = (18.75, 0, 0) are obtained by interpolation calculation. L ^* a ^* b ^* value = (12.5, 0, 0) and L ^* a ^* b ^* value = (25, 0, 0) are selected as L ^* a ^* b ^* values used in the interpolation calculation. Shall be. Here, by referring to the color conversion table of FIG. 3, (273, 271, 257) is obtained as the CMY value corresponding to L ^* a ^* b ^* value = (12.5, 0, 0), and L ^* (228, 224, 208) is obtained as the CMY value corresponding to a ^* b ^* value = (25, 0, 0).

And L ^* a ^* b ^* value = (18.75,0,0) is L ^* a ^* b ^* value = (12.5,0,0) and L on the L ^* a ^* b ^* space. ^Since this is in the middle of ^* a ^* b ^* value = (25, 0, 0), this relationship is applied to the CMY space, and CMY value = (273, 271, 257) and CMY value = (228, 224). , 208) is a CMY value corresponding to L ^* a ^* b ^* value = (18.75, 0, 0), CMY = (250.5, 247.5, 232.5) Thus, it is possible to calculate a value that is very close to the actual CMY value = (252, 253, 235).

  Next, a method for generating a color signal that exceeds the color reproduction range of the device will be described. When registering the correspondence relationship of the color signal exceeding the color reproduction range of the device in the color conversion table, the color signal exceeding the color reproduction range of the device cannot be actually reproduced. Therefore, the correspondence between the color signals in the color gamut is obtained, and the correspondence between the color signals in the color gamut also holds between the color signals outside the color gamut. By expanding outside the range, a color signal exceeding the color reproduction range of the device is generated.

FIG. 4 is a block diagram showing a configuration of a color signal generation apparatus according to an embodiment of the present invention.
In FIG. 4, the color signal generation device 11 includes a color signal selection unit 12 and a color signal extrapolation unit 13. The color signal selection unit 12 selects the correspondence between the color signals in the device color gamut, and the color signal extrapolation unit 13 extrapolates the correspondence between the color signals in the device color gamut. A color signal that exceeds the color reproduction range of the device is generated.

  In this way, by extrapolating the correspondence between the color signals in the device color gamut, it is possible to calculate the color signal for a color that cannot be measured, and the color signal outside the color reproduction range of the device can be calculated. Can be generated.

FIG. 5 is a flowchart showing a color signal generation method according to an embodiment of the present invention.
In FIG. 5, first, a correspondence relationship between color signals in the device color gamut is acquired (step S1), and a correspondence relationship outside the device color gamut is generated by using the color signal correspondence relationship in the device color gamut (step S1). Step S2).

FIG. 6 is a block diagram showing a system configuration of a color signal generation apparatus or color conversion apparatus according to an embodiment of the present invention.
In FIG. 6, 21 is a central processing unit (CPU) that performs overall processing, 22 is a read only memory (ROM), 23 is a random access memory (RAM), 24 is an input / output interface, and 25 is based on RGB signals. A display for displaying a color image; 26, a printer for printing a color image based on the CMY signal; 27, a memory for temporarily storing data read by the scanner 28; and 28, an RGB signal by reading the color image. , A communication interface, 30 a communication network, 31 a driver for driving a storage medium, 32 a hard disk, 33 an IC memory card, 34 a magnetic tape, 35 a flexible disk, 36 a CD-ROM, An optical disk such as a DVD-ROM, 37 is a bar , 38 is a keyboard.

  A program for performing color signal generation processing, a program for performing color signal conversion processing, or the content of a color conversion table is stored in a storage medium such as the hard disk 32, IC memory card 33, magnetic tape 34, flexible disk 35, optical disk 36, etc. By reading the color signal generation process program, the color signal conversion process program, or the contents of the color conversion table into the RAM 23, the color signal generation process and the color signal conversion process can be performed. Also, the ROM 22 can store a program for performing color signal generation processing, a program for performing color signal conversion processing, or the contents of a color conversion table.

  Furthermore, the program for performing color signal generation processing, the program for performing color signal conversion processing, or the contents of the color conversion table can be extracted from the communication network 30 via the communication interface 29. The communication network 30 connected to the communication interface 29 includes, for example, a LAN (Local Area Network), a WAN (Wide Area Network), the Internet, an analog telephone network, a digital telephone network (ISDN: Integrated Service Digital Network), and a PHS (Personal Handy Network). System) and satellite communication.

The CPU 21 sends a CMY signal to the printer 26 when a program for performing color signal generation processing is started. When the CMY signal is sent, the printer 26 prints a color corresponding to the CMY signal. This color is read by a colorimeter, and an L ^* a ^* b ^* value corresponding to the color is input from the keyboard 38. The CPU 21 extrapolates the correspondence relationship between the CMY signal sent to the printer 26 and the L ^* a ^* b ^* value input from the keyboard 38, so that the CMY signal exceeding the color reproduction range of the printer 26 and the L ^* a A correspondence relationship with ^* b ^* value is generated and registered in the color conversion table.

The CPU 21 receives the RGB signal read by the scanner 28 when a program for performing color conversion processing is started. When the RGB signal read by the scanner 28 is received, the RGB signal is converted into an L ^* a ^* b ^* signal by referring to the color conversion table of the scanner 28. The L ^* a ^* b ^* signal is converted into a CMY signal by referring to the color conversion table of the printer 26 and output to the printer 26.

Next, a description will be given of a first embodiment of a method for creating a conversion table for converting L ^* a ^* b ^* values into CMY values.
FIG. 7 is a diagram showing a color signal generation process in the color space according to the first embodiment of the color conversion table creation method. In the first embodiment, extrapolation is performed by using a correspondence relationship between L ^* a ^* b ^* values having a lattice-like regular distribution and CMY values having an irregular distribution. A correspondence relationship between CMY values outside the printer color gamut and L ^* a ^* b ^* values is generated.

  In FIG. 7, colors regularly distributed as CMY values are printed by a printer. The colors printed by this printer have a regular grid-like distribution in the CMY space and exist only within the color reproduction range of the printer.

Next, the color measurement is performed on the result of printing by this printer. The colors measured by this colorimeter are distributed in the L ^* a ^* b ^* space, and the distribution of L ^* a ^* b ^* values corresponding to the CMY values of the grid points in the CMY space is irregular. That is, correspondence between the L ^* a ^* b ^* values distributed irregularly in CMY values and L ^* a ^* b ^* space which regularly distributed on the CMY space is obtained.

Next, the correspondence between the L ^* a ^* b ^* values distributed irregularly in CMY values and L ^* a ^* b ^* space which regularly distributed over the CMY space, L ^* a ^* b ^* space To make a regular distribution. This transformation, L ^* a ^* b ^* performs interpolation of L ^* a ^* b ^* values irregularly distributed in space, L ^* a ^* b ^* values at lattice points in space L ^* a ^* b ^* value This can be done by obtaining the correspondence between CMY values and CMY values.

As this interpolation calculation, there is a method in which a plurality of set CMY values are selected so as to surround unset CMY values, and a weight calculation is performed on L ^* a ^* b ^* values corresponding to the CMY values. Then, among the L ^* a ^* b ^* values obtained by interpolation, the one on the lattice point in the L ^* a ^* b ^* space is selected, and the CMY corresponding to the L ^* a ^* b ^* value is selected. Register the value in the color conversion table.

Next, when the correspondence between the L ^* a ^* b ^* values regularly distributed in the L ^* a ^* b ^* space and the CMY values in the printer gamut is obtained, the L ^* a ^* b in the printer gamut is obtained. By extrapolating the correspondence between the ^* value and the CMY value, the correspondence between the CMY value outside the printer color gamut and the L ^* a ^* b ^* value is generated.

As described above, according to the first embodiment of the color conversion table creation method, the color reproduction range of the device is obtained by extrapolating the color signals of the lattice points in the L ^* a ^* b ^* space whose color reproducibility does not depend on the device. By generating the correspondence between external color signals, extrapolation points in the L ^* a ^* b ^* space can be generated at uniform intervals, allowing efficient color conversion beyond the color reproduction range of the device. It is possible to generate a color conversion table that can be performed at high speed.

Hereinafter, the relationship between the L ^* a ^* b ^* values irregularly distributed on the CMY values and L ^* a ^* b ^* space regularly distributed over the CMY space, in the L ^* a ^* b ^* space A method for generating a correspondence relationship between regularly distributed L ^* a ^* b ^* values and CMY values will be described.

8, the correspondence between the correspondence between the CMY values and L ^* a ^* b ^* value that is set on the grid points in the CMY space, the CMY values and L ^* a ^* b ^* values between the grid points It is a figure explaining the method of interpolating.

FIG. 8A shows an interpolation method using cubic interpolation, and the correspondence between CMY values and L ^* a ^* b ^* values distributed in a discrete manner on lattice points in the CMY space is set in advance. To do. Here, assuming that the input CMY value is at the position of the point O in the CMY space, a unit cube that includes the point O exists in the CMY space, and the eight vertices P0 that constitute the unit cube are included. L ^* a ^* b ^* values set corresponding to .about.P7 are obtained.

Further, if the unit cube is divided by a plane passing through the point O and perpendicular to each axis of CMY, eight rectangular parallelepipeds are formed. When the volume of this rectangular parallelepiped is V0 to V7 corresponding to the eight vertices P0 to P7, and the L ^* a ^* b ^* value corresponding to the position of the point O is Oout,
Oout = (P0 · V7 + P1 · V6 + P2 · V5 + P3 · V4
+ P4 ・ V3 + P5 ・ V2 + P6 ・ V1 + P7 ・ V0) / VT
(However, VT = V0 + V1 + V2 + V3 + V4 + V5 + V6 + V7)
... (1)
It is obtained by.

Note that P0 to P7 in (1) are L ^* a ^* b ^* values set corresponding to the eight vertices P0 to P7 in the CMY space.
FIG. 8B shows an interpolation method using triangular prism interpolation. P0 to P7 are eight vertices of a unit cube including a point O therein, and DL, Da, and Db are standards determined by the point O. This is the distance in each axial direction. Here, in the triangular prism interpolation, the unit cube is divided into two regions, and to which point the point O belongs can be determined by evaluating whether Da ≧ Db. If Da ≧ Db is true, the point O belongs to the triangular prism E2, and if Da ≧ Db is false, the point O belongs to the triangular prism E1.

When the point O belongs to the triangular prism E2, Oout corresponding to the position of the point O is
m = P5 + Da · (P4−P5) + Db · (P7−P4)
n = P0 + Da · (P1−P0) + Db · (P2−P1)
Oout = m + DL. (Nm) (2)
It can ask for.

FIG. 8C shows an interpolation method by tetrahedral interpolation. P0 to P7 are eight vertices of a unit cube including a point O inside, and DL, Da, and Db are determined by the point O. This is the normalized distance in each axial direction. Here, in the tetrahedral interpolation, the unit cube is divided into six tetrahedrons, and the point O belongs to can be determined by evaluating DL ≧ Da, Da> Db, and Db> DL. That is,
If DL ≧ Da is true, Da> Db is true, and Db> DL is false, the point O belongs to the tetrahedron P5P0P1P2,
If DL ≧ Da is true, Da> Db is false, and Db> DL is false, the point O belongs to the tetrahedron P5P0P3P2,
If DL ≧ Da is true, Da> Db is false, and Db> DL is true, the point O belongs to the tetrahedron P5P6P3P2,
If DL ≧ Da is false, Da> Db is true, and Db> DL is true, the point O belongs to the tetrahedron P5P4P7P2,
If DL ≧ Da is false, Da> Db is true, and Db> DL is false, the point O belongs to the tetrahedron P5P6P7P2,
If DL ≧ Da is false, Da> Db is false, and Db> DL is true, the point O belongs to the tetrahedron P5P4P1P2.

When the point O belongs to the tetrahedron P5P0P1P2, Oout corresponding to the position of the point O is
Oout = P5 + DL. (P0-P5) + Da. (P1-P0)
+ Db. (P2-P1) (3)
It can ask for.

FIG. 9 is a diagram for two-dimensionally explaining a method for obtaining a correspondence relationship between L ^* a ^* b ^* values regularly distributed in the L ^* a ^* b ^* space and CMY values by interpolation.
In FIG. 9, a color corresponding to a grid point in the CMY space is printed by a printer, and the color printed by the printer is measured by a colorimeter, whereby the CMY value and L ^* a ^* of the grid point in the CMY space are measured ^. It is assumed that the correspondence with the L ^* a ^* b ^* values in the b ^* space is generated discretely. Here, the L ^* a ^* b ^* values corresponding to the CMY values of the lattice points in the CMY space are irregularly distributed in the L ^* a ^* b ^* space. Thus, by interpolating the correspondence relationship between the L ^* a ^* b ^* values of the CMY values of grid points in the CMY space and L ^* a ^* b ^* space, grid points in the L ^* a ^* b ^* space M1~M25 CMY values corresponding to the L ^* a ^* b ^* values are generated.

Specifically, for example, a square composed of four grid points for which CMY values are set is divided into 4 × 4 squares (in an actual three-dimensional CMY space, CMY values are set). A cube composed of 8 lattice points is divided into 4 × 4 × 4 cubes), and L ^* a ^* b ^* values corresponding to CMY values at each divided lattice point are obtained by interpolation. Then, among the calculated by interpolation L ^* a ^* b ^* values, select the one closest to the L ^* a ^* b ^* value at the grid point M1~M25 the L ^* a ^* b ^* space, and the selected L ^* a ^* b ^* values and the corresponding relationship between the CMY value is registered as the corresponding relationship between the CMY values of the L ^* a ^* b ^* grid points of space L ^* a ^* b ^* values and CMY space.

For example, it is assumed that the points P1 to P19 in the CMY space correspond to the points Q1 to Q19 in the L ^* a ^* b ^* space. Then, L ^* a ^* b ^* of being set L ^* a ^* b ^* values in space, L ^* a ^* b ^* values of the L ^* a ^* b ^* nearest the point of the value Q1 of the grid point M1 In this case, the CMY value of the point P1 in the CMY space is registered as the CMY value corresponding to the L ^* a ^* b ^* value of the lattice point M1. Further, L ^* a ^* b ^* of the L ^* a ^* b ^* values that are set in the space, the closest to the L ^* a ^* b ^* values of grid points M2 is the point Q2 L ^* a ^* b ^* value In this case, the CMY value of the point P2 in the CMY space is registered as the CMY value corresponding to the L ^* a ^* b ^* value of the lattice point M2. Similarly, the L ^* a ^* b ^* value of the point Q3 is selected as the closest to the L ^* a ^* b ^* value of the lattice point M3, thereby corresponding to the L ^* a ^* b ^* value of the lattice point M3. By registering the CMY value of the point P3 in the CMY space as the CMY value to be performed and selecting the L ^* a ^* b ^* value of the point Q4 as being closest to the L ^* a ^* b ^* value of the grid point M4, The CMY value of the point P4 in the CMY space is registered as the CMY value corresponding to the L ^* a ^* b ^* value of the grid point M4, and the L of the point Q5 is assumed to be the closest to the L ^* a ^* b ^* value of the grid point M5. ^* by a ^* b ^* value is selected, as the CMY values corresponding to L ^* a ^* b ^* values of grid points M5, CMY value of the point in the CMY space P5 is registered.

Further, the CMY value of the point P9 in the CMY space is registered corresponding to the L ^* a ^* b ^* value of the lattice point M6, and the point P6 of the CMY space is corresponded to the L ^* a ^* b ^* value of the lattice point M7. CMY value is registered, the CMY values of L ^* a ^* b ^* point of CMY space corresponding to the value P7 of the lattice points M8 is registered, the CMY space corresponding to the L ^* a ^* b ^* values of grid points M9 the CMY value of the point P8 is registered, CMY values of L ^* a ^* b ^* point of CMY space corresponding to the value P15 of the lattice points M11 is registered, corresponding to the L ^* a ^* b ^* values of grid points M12 CMY value of the point P12 in the CMY space is registered, CMY value of the point P10 of the lattice points M13 L ^* a ^* b ^* values in the corresponding CMY space is registered, the grid points M14 L ^* a ^* b ^* CMY value of the point in the CMY space P11 corresponding to the value is registered, the grid points M17 L ^* a ^* b ^* values in the corresponding CM CMY values at the point of space P16 is registered, the CMY values of L ^* a ^* b ^* point of CMY space corresponding to the value P14 of the lattice points M18 is registered, the L ^* a ^* b ^* values of grid points M19 corresponding CMY value of the point P13 in the CMY space is registered, the CMY values of the L ^* a ^* b ^* point of CMY space corresponding to the value P17 of the lattice points M22 is registered, the grid points M23 L ^* a ^* CMY value of the point in the CMY space P18 corresponding to the b ^* values are registered, CMY values of L ^* a ^* b ^* point of CMY space corresponding to the value P19 of the lattice points M24 is registered.

In this way, the correspondence relationship of the color signals is increased by interpolation on the regularly distributed CMY space, and from this, the one close to the lattice point on the L ^* a ^* b ^* space is found, whereby L ^* a ^From the correspondence between the color signal irregularly distributed in the ^* b ^* space and the color signal regularly distributed in the CMY space, it is distributed to the lattice points in the L ^* a ^* b ^* space. It is possible to approximately obtain the correspondence between the color signal and the color signal in the CMY space, and by reducing the division interval in the CMY space, the lattice point in the L ^* a ^* b ^* space can be obtained. It is possible to improve the accuracy of the correspondence between the distributed color signals and the color signals in the CMY space.

FIG. 10 is a diagram showing a configuration for converting L ^* a ^* b ^* values that are irregularly distributed in the L ^* a ^* b ^* space into L ^* a ^* b ^* values that are regularly distributed.
In FIG. 10, the printer 41 can output CMY256 gradation data, and prints a color corresponding to the CMY value by designating each CMY value with a value of 0 to 255. The colorimeter 42 measures the color chart printed by the printer 41 and outputs an L ^* a ^* b ^* value corresponding to the color of the color chart.

When the CMY number 44 is designated from the color conversion table 43, the CMY value calculation unit 46 calculates a CMY value corresponding to the CMY number 44 and outputs it to the printer 41. The printer 41 prints a color corresponding to the CMY value, and the colorimeter 42 measures the color chart printed by the printer 41, whereby L ^* a ^* b corresponding to the color of the color chart. ^* Output the value. The L ^* a ^* b ^* value 45 output by the colorimeter 42 is registered in the color conversion table 43.

Here, assuming that an L ^* a ^* b ^* value 45 corresponding to a CMY value distributed in a grid pattern in the CMY space and changed to nine levels is registered in the color conversion table 43, the CMY number 44 is varied between 0 and 8, respectively. In this case, the CMY value corresponding to the CMY number 44 is
CMY value = (int) (255.0 × CMY number / 8.0 + 0.5)
(However, int is rounded down) (4)
It can ask for.

In the color conversion table 43, as is apparent from the equation (4), the CMY values in the CMY space have a grid-like distribution. However, L ^* a ^* b ^* values in the L ^* a ^* b ^* space, as shown in FIG. 9, has become an irregular distribution. Therefore, a correspondence relationship between the L ^* a ^* b ^* values and the CMY values in which the L ^* a ^* b ^* values in the L ^* a ^* b ^* space have a grid-like distribution is generated, and the color conversion table 50 has sign up.

Specifically, the interpolation unit 47 performs interpolation of the color conversion table 43 by using the correspondence relationship between the CMY number 44 registered in the color conversion table 43 and the L ^* a ^* b ^* value 45, and CMY. Increase the correspondence between the value and the L ^* a ^* b ^* value. For example, a unit cube generated by eight grid points in the CMY space is divided into 4 × 4 × 4, and L ^* a ^* b ^* values corresponding to the CMY values of the division points are divided into eight pieces surrounding the division points. It is obtained by weight calculation of L ^* a ^* b ^* values corresponding to lattice points.

Next, the proximity color detection unit 48 detects a lattice point having an L ^* a ^* b ^* value close to the L ^* a ^* b ^* value obtained by the interpolation unit 47 from the L ^* a ^* b ^* space, and this lattice. The L ^* a ^* b ^* number corresponding to the point is calculated.

Then, the proximity calculating unit 49 calculates the proximity between the L ^* a ^* b ^* values of grid points detected by the proximity color detection unit 48 and the L ^* a ^* b ^* value obtained by the interpolation section 47 . Note that proximity is that defined by the square of the distance between the L ^* a ^* b ^* values of grid points detected by the proximity color detection unit 48 and the L ^* a ^* b ^* value obtained by the interpolation section 47 Can do.

Next, the comparison / selection unit 50 reads the proximity 55 of the L ^* a ^* b ^* number 53 calculated by the proximity color detection unit 48 from the color conversion table 52, and the proximity 55 read from the color conversion table 52, The proximity calculated by the proximity calculation unit 49 is compared. If the proximity 55 read from the color conversion table 52 is greater than the proximity calculated by the proximity calculation unit 49, the L ^* a ^* b ^* number 53 calculated by the proximity color detection unit 48 and the proximity are calculated. The degree of proximity calculated by the degree calculation unit 49 is output to the registration unit 51.

Next, the registration unit 51 correlates with the L ^* a ^* b ^* number 53 output from the comparison / selection unit 50, the CMY value output from the interpolation unit 47, and the proximity level output from the comparison / selection unit 50. Are registered in the color conversion table 52.

The range of L ^* a ^* b ^* values registered in the color conversion table 52 is, for example, 0 ≦ L ^* ≦ 100, −100 ≦ a ^* ≦ 100, −100 ≦ b ^* including the color reproduction range of the printer 41 ^. It is assumed that ≦ 100, and L ^* a ^* b ^* values in the L ^* a ^* b ^* space are distributed in a grid pattern at 16 intervals. Therefore, L ^* a ^* b ^* numbers, 0 ≦ L ^* number ≦ 7,0 ≦ a ^* number ≦ 13, 0 ≦ b ^* number ≦ 13 next, L ^* a ^* b ^* numbers and L ^* a ^* b ^* The relationship with the value is
L ^* value = L ^* number × 16 (5)
a ^* value = a ^* number × 16−100 (6)
b ^* value = b ^* number × 16−100 (7)
Is required.

  FIG. 11 is a flowchart showing a process for selecting a set point of CMY values used for interpolating a non-set point of CMY values. In the flowchart of FIG. 11, CMY values are registered in the color conversion table 43 in the range of 0 ≦ C number ≦ 8, 0 ≦ M number ≦ 8, and 0 ≦ Y number ≦ 8.

In FIG. 11, first, y = 0, m = 0, and c = 0 are set (step S11). Here, y represents the Y number, m represents the M number, and c represents the C number.
Next, as eight lattice points P0 to P7 constituting a unit cube of the CMY space,
P0 = P (y, m, c) (8)
P1 = P (y + 1, m, c) (9)
P2 = P (y, m + 1, c) (10)
P3 = P (y, m, c + 1) (11)
P4 = P (y + 1, m + 1, c) (12)
P5 = P (y, m + 1, c + 1) (13)
P6 = P (y + 1, m, c + 1) (14)
P7 = P (y + 1, m + 1, c + 1) (15)
Is selected (step S12). Here, P (y, m, c) indicates the position of the lattice point indicated by y, m, c.

Next, by dividing the L ^* a ^* b ^* values corresponding to the eight selected grid points P0 to P7, the division points in the unit cube designated by the eight grid points P0 to P7 are calculated. corresponding to the L ^* a ^* b ^* calculated values, L ^* a ^* b ^* grid points of space L ^* a ^* b ^* is closest to the value registered in the color conversion table 52 (step S13).

  Hereinafter, by incrementing y, m, or c by 1, 8 × 8 × 8 unit cubes registered in the color conversion table 43 are sequentially selected, and the interpolation calculation is repeated (steps S14 to S19). ).

FIG. 12 is a flowchart showing the interpolation calculation process in step S13 of FIG. In the flowchart of FIG. 12, the unit cube designated by the eight lattice points P0 to P7 is divided into 4 × 4 × 4 and the range of L ^* a ^* b ^* values registered in the color conversion table 52. Is set to 0 ≦ L ^* ≦ 100, −100 ≦ a ^* ≦ 100, −100 ≦ b ^* ≦ 100, and L ^* a ^* b ^* values in the L ^* a ^* b ^* space are distributed in a lattice pattern at 16 intervals. It shows the case of generating things.

  In FIG. 12, first, i = 0, j = 0, k = 0, and n = 4 are set (step S21). Here, i, j, and k are variables indicating the positions of the division points of the unit cube. Since the unit cube is divided into 4 × 4 × 4, 0 ≦ i ≦ 3, 0 ≦ j ≦ 3, 0 ≦ k ≦ 3.

Next, for the division point designated by i, j, k, a weight for obtaining the L ^* a ^* b ^* value corresponding to this division point is generated. Here, it is based on cube interpolation, and in cube interpolation, the volumes of eight rectangular parallelepipeds that are divided by planes that pass through the dividing points and intersect perpendicularly to the respective CMY axes are weights used in the interpolation calculation. That is, the weights W0 to W7 are
W0 = (n−i) × (n−j) × (n−k) (16)
W1 = i × (n−j) × (n−k) (17)
W2 = (n−i) × j × (n−k) (18)
W3 = (n−i) × (n−j) × k (19)
W4 = i × j × (n−k) (20)
W5 = (n−i) × j × k (21)
W6 = i × (n−j) × k (22)
W7 = i × j × k (23)
It becomes.

By using these weights W0 to W7, the L ^* a ^* b ^* values corresponding to the division points are
L ^* = (P0 (L ^* ) × W0 + P1 (L ^* ) × W1 + P2 (L ^* ) × W2
+ P3 (L ^* ) × W3 + P4 (L ^* ) × W4 + P5 (L ^* ) × W5
P6 (L ^* ) × W6 + P7 (L ^* ) × W7) / n3 (24)
a ^* = (P0 (a ^* ) × W0 + P1 (a ^* ) × W1 + P2 (a ^* ) × W2
+ P3 (a ^* ) × W3 + P4 (a ^* ) × W4 + P5 (a ^* ) × W5
P6 (a ^* ) × W6 + P7 (a ^* ) × W7) / n3 (25)
b ^* = (P0 (b ^* ) × W0 + P1 (b ^* ) × W1 + P2 (b ^* ) × W2
+ P3 (b ^* ) × W3 + P4 (b ^* ) × W4 + P5 (b ^* ) × W5
P6 (b ^* ) × W6 + P7 (b ^* ) × W7) / n3 (26)
(Step S22). However, P0 (L ^* ) to P7 (L ^* ) are L ^* values corresponding to the lattice points P0 to P7 selected in FIG. 11, and P0 (a ^* ) to P7 (a ^* ) are selected in FIG. The a ^* values P0 (b ^* ) to P7 (b ^* ) corresponding to the lattice points P0 to P7 thus set are the b ^* values corresponding to the lattice points P0 to P7 selected in FIG.

Also, the CMY values of the dividing points specified by i, j, k are
Y = 255.0 × (y + i / n) /8.0 (27)
M = 255.0 × (m + j / n) /8.0 (28)
C = 255.0 × (c + k / n) /8.0 (29)
It is. However, y, m, and c are variables indicating the position of the unit cube.

Then, the obtained L ^* a ^* b ^* values, by calculating whether exists between any grid point in the L ^* a ^* b ^* space, close to the L ^* a ^* b ^* value obtained Lattice points in the L ^* a ^* b ^* space are detected (step S23). That, L ^* a ^* b ^* number of grid points in the L ^* a ^* b ^* space adjacent to the L ^* a ^* b ^* value obtained is
L ^* number = (int) ((L ^* + 8) / 16) (30)
a ^* number = (int) ((a ^* + 108) / 16) (31)
b ^* number = (int) ((b ^* + 108) / 16) (32)
It is obtained by. However, (int) represents truncation after the decimal point.

Next, the degree of proximity D between the obtained L ^* a ^* b ^* value and the lattice point in the L ^* a ^* b ^* space is expressed as follows ^:
D = (L ^* number × 16−L) ² + (a ^* number × 16−100−a) ²
+ (B ^* number × 16−100−b) ² (33)
(Step S24).

Next, when the CMY value corresponding to the L ^* a ^* b ^* number is not registered in the color conversion table 52, or the proximity degree corresponding to the L ^* a ^* b ^* number already registered in the color conversion table 52 Is greater than the proximity obtained in step S24 (step S25), the CMY value obtained in step S22 and the proximity obtained in step S24 in correspondence with the L ^* a ^* b ^* number obtained in step S23. Are registered in the color conversion table 52 (step S26).

  Hereinafter, by increasing i, j, or k by 1, the division points of the unit cube divided into 4 × 4 × 4 are sequentially selected, and the interpolation calculation and the update process of the color conversion table 52 are repeated ( Step S27 to Step S32).

As described above, the correspondence between the L ^* a ^* b ^* values regularly distributed in the L ^* a ^* b ^* space and the CMY values in the color gamut of the printer 41 is generated.
Next, as shown in FIG. 7, extrapolating the correspondence between the L ^* a ^* b ^* values regularly distributed in the L ^* a ^* b ^* space and the CMY values in the color gamut of the printer 41. Thus, a correspondence relationship between L ^* a ^* b ^* values regularly distributed in the L ^* a ^* b ^* space and CMY values outside the color gamut in the printer 41 is generated.

FIG. 13 is a diagram showing the distribution of grid points inside and outside the color gamut of the printer on the a ^* b ^* plane of the L ^* a ^* b ^* space.
In FIG. 13, assuming that the CMY values corresponding to the grid points (black circles) in the printer color gamut are obtained by the processing of FIGS. 11 and 12, the CMY values correspond to the grid points (white circles) outside the printer color gamut. The CMY value can be obtained by extrapolating CMY values corresponding to grid points (black circles) in the printer color gamut. For example, when the CMY value corresponding to the grid point P is obtained, the grid point P1 adjacent to the grid point P and the grid point P2 adjacent to the grid point P1 are set to a grid inside the boundary A in the color gamut of the printer. Take out from the point. Then, the CMY value corresponding to the grid point P is calculated by using the CMY value corresponding to the grid point P1 and the CMY value corresponding to the grid point P2. If the grid point P1 adjacent to the grid point P and the grid point P2 adjacent to the grid point P1 are not adjacent to each other in the color gamut of the printer, the processing is moved to the next grid point in the L ^* a ^* b ^* space. .

By performing the above processing on all grid points in the L ^* a ^* b ^* space, CMY values can be set at grid points inside the boundary B around the printer color gamut. Next, by using the data of the grid points inside the boundary B, CMY values are set at the grid points inside the boundary C, and further, using the data of the grid points inside the boundary C, CMY values can be set at inner grid points.

FIG. 14 is a diagram for explaining a method for calculating CMY values corresponding to the grid points P in FIG.
In FIG. 14, the lattice point P1 is adjacent to the lattice point P, and the lattice point P2 is adjacent to the lattice point P. The grid point P1 corresponds to the point Q1 in the CMY space, the grid point P2 corresponds to the point Q2 in the CMY space, and the CMY value of the point Q1 and the CMY value of the grid point Q2 are registered in the color conversion table. Shall. At this time, the CMY value corresponding to the lattice point P can be obtained as the CMY value of the point Q that is symmetric with respect to the point Q2 with respect to the point Q1.

Hereinafter, it demonstrates more concretely using numerical formula. In the color conversion table to be generated, the L ^* a ^* b ^* values are 16 intervals each, and the range of the L ^* a ^* b ^* values is different from the actual color existence area, but 0 ≦ L ^* ≦ 112, −128 ≦ a ^* ≦ 128, −128 ≦ b ^* ≦ 128. L ^* a ^* b ^* values are managed by L ^* a ^* b ^* numbers (0 ≦ Lnum ≦ 7, anum ≦ 17, 0 ≦ bnum ≦ 17). For example, L ^* a ^* b ^* number = (0, 0,0) corresponds to a lattice point of L ^* a ^* b ^* value = (0, −128, −128), and L ^* a ^* b ^* number = (0,0,1) is L ^* a Corresponding to the lattice point of ^* b ^* value = (0, −128, −112), L ^* a ^* b ^* number = (5, 17, 17) is L ^* a ^* b ^* value = (112,128) , 128). That is,
L ^* = Lnum · 16 (34)
a ^* = anum · 16−128 (35)
b ^* = bnum · 16−128 (36)
It is. However, Lnum represents an L ^* number, annum represents an a ^* number, and bnum represents a b ^* number.

The actual storage format of the color conversion table is a three-dimensional array in which each of the C value, the M value, and the Y value has an L ^* a ^* b ^* number as an argument. That is,
C [Lnum] [anum] [bnum],
M [Lnum] [anum] [bnum],
Y [Lnum] [anum] [bnum]
And

Then, all grid points are sequentially selected, and the subsequent processing is performed for each grid point. However, the L ^* a ^* b ^* number of the selected lattice point is (Lnum, anum, bnum).
First, it is investigated whether data is set at the selected grid point. In the initial state, data is set only for grid points in the color gamut. Then, a grid point for which no data is set is selected, and it is checked whether data for a grid point adjacent to the selected grid point and data for a further adjacent grid point have been set. Here, the L ^* a ^* b ^* number of the lattice point adjacent to the selected lattice point is set to (Lnum1, annum1, bnum1), and the L ^* a ^* b ^* number of the adjacent lattice point is set to (Lnum2, annum2, bnum2). ).

FIG. 15 is a diagram illustrating positions of lattice points K1 to K26 adjacent to the selected lattice point O. As illustrated in FIG. As shown in FIG. 15, there are only 26 grid points K1 to K26 adjacent to the selected grid point O. L ^* a ^* b ^* number of grid points K1~K26 the next selected grid point O (Lnum1, anum1, bnum1) , and the lattice points of the further next selected grid points O L ^* a ^* b ^* number ( When Lnum2, annum2, bnum2) are expressed using L ^* a ^* b ^* numbers (Lnum, anum, bnum) of the selected lattice point O, the following equation is obtained.
1) (Lnum1, anum1, bnum1)
= (Lnum + 1, anum, bnum) (K1)
(Lnum2, anum2, bnum2)
= (Lnum + 2, anum, bnum)
2) (Lnum1, anum1, bnum1)
= (Lnum-1, anum, bnum) (K2)
(Lnum2, anum2, bnum2)
= (Lnum-2, anum, bnum)
3) (Lnum1, anum1, bnum1)
= (Lnum, anum + 1, bnum) (K3)
(Lnum2, anum2, bnum2)
= (Lnum, anum + 2, bnum)
4) (Lnum1, anum1, bnum1)
= (Lnum, anum-1, bnum) (K4)
(Lnum2, anum2, bnum2)
= (Lnum, anum-2, bnum)
5) (Lnum1, anum1, bnum1)
= (Lnum, anum, bnum + 1) (K5)
(Lnum2, anum2, bnum2)
= (Lnum, anum, bnum + 2)
6) (Lnum1, anum1, bnum1)
= (Lnum, anum, bnum-1) (K6)
(Lnum2, anum2, bnum2)
= (Lnum, anum, bnum-2)
7) (Lnum1, anum1, bnum1)
= (Lnum + 1, anum + 1, bnum) (K7)
(Lnum2, anum2, bnum2)
= (Lnum + 2, anum + 2, bnum)
8) (Lnum1, anum1, bnum1)
= (Lnum + 1, anum-1, bnum) (K8)
(Lnum2, anum2, bnum2)
= (Lnum + 2, anum-2, bnum)
9) (Lnum1, anum1, bnum1)
= (Lnum-1, anum + 1, bnum) (K9)
(Lnum2, anum2, bnum2)
= (Lnum-2, anum + 2, bnum)
10) (Lnum1, anum1, bnum1)
= (Lnum-1, anum-1, bnum) (K10)
(Lnum2, anum2, bnum2)
= (Lnum-2, anum-2, bnum)
11) (Lnum1, anum1, bnum1)
= (Lnum + 1, anum, bnum + 1) (K11)
(Lnum2, anum2, bnum2)
= (Lnum + 2, anum, bnum + 2)
12) (Lnum1, anum1, bnum1)
= (Lnum + 1, anum, bnum-1) (K12)
(Lnum2, anum2, bnum2)
= (Lnum-2, anum, bnum-2)
13) (Lnum1, anum1, bnum1)
= (Lnum-1, anum, bnum + 1) (K13)
(Lnum2, anum2, bnum2)
= (Lnum-2, anum, bnum + 2)
14) (Lnum1, anum1, bnum1)
= (Lnum-1, anum, bnum-1) (K14)
(Lnum2, anum2, bnum2)
= (Lnum-2, anum, bnum-2)
15) (Lnum1, anum1, bnum1)
= (Lnum, anum + 1, bnum + 1) (K15)
(Lnum2, anum2, bnum2)
= (Lnum, anum + 2, bnum + 2)
16) (Lnum1, anum1, bnum1)
= (Lnum, anum + 1, bnum-1) (K16)
(Lnum2, anum2, bnum2)
= (Lnum, anum + 2, bnum-2)
17) (Lnum1, anum1, bnum1)
= (Lnum, anum-1, bnum + 1) (K17)
(Lnum2, anum2, bnum2)
= (Lnum, anum-2, bnum + 2)
18) (Lnum1, anum1, bnum1)
= (Lnum, anum-1, bnum-1) (K18)
(Lnum2, anum2, bnum2)
= (Lnum, anum-2, bnum-2)
19) (Lnum1, anum1, bnum1)
= (Lnum + 1, anum + 1, bnum + 1) (K19)
(Lnum2, anum2, bnum2)
= (Lnum + 2, anum + 2, bnum + 2)
20) (Lnum1, anum1, bnum1)
= (Lnum + 1, anum + 1, bnum-1) (K20)
(Lnum2, anum2, bnum2)
= (Lnum + 2, anum + 2, bnum-2)
21) (Lnum1, anum1, bnum1)
= (Lnum + 1, anum-1, bnum + 1) (K21)
(Lnum2, anum2, bnum2)
= (Lnum + 2, anum-2, bnum + 2)
22) (Lnum1, anum1, bnum1)
= (Lnum + 1, anum-1, bnum-1) (K22)
(Lnum2, anum2, bnum2)
= (Lnum + 2, anum-2, bnum-2)
23) (Lnum1, anum1, bnum1)
= (Lnum-1, anum + 1, bnum + 1) (K23)
(Lnum2, anum2, bnum2)
= (Lnum-2, anum + 2, bnum + 2)
24) (Lnum1, anum1, bnum1)
= (Lnum-1, anum + 1, bnum-1) (K24)
(Lnum2, anum2, bnum2)
= (Lnum-2, anum + 2, bnum-2)
25) (Lnum1, anum1, bnum1)
= (Lnum-1, anum-1, bnum + 1) (K25)
(Lnum2, anum2, bnum2)
= (Lnum-2, anum-2, bnum + 2)
26) (Lnum1, anum1, bnum1)
= (Lnum-1, anum-1, bnum-1) (K26)
(Lnum2, anum2, bnum2)
= (Lnum-2, anum-2, bnum-2)
As described above, since there are 26 directions corresponding to the next to the selected lattice point O, it is sequentially examined in the order of 1) to 26). This order is an example, and other orders may be used. When a grid point adjacent to the selected grid point and a data set with a grid point adjacent to the selected grid point are found, the process proceeds to extrapolation calculation processing.

Extrapolation, especially in CMY values corresponding to the grid points in the L ^* a ^* b ^* number (Lnum1, anum1, bnum1), corresponding to the grid points in the L ^* a ^* b ^* number (Lnum2, anum2, bnum2) In the process of obtaining a CMY value that is symmetrical to the CMY value to be calculated, it can be calculated by the following equation.

C [Lnum] [anum] [bnum]
= C [Lnum1] [anum1] [bnum1] · 2
-C [Lnum2] [anum2] [bnum2] (37)
M [Lnum] [anum] [bnum]
= M [Lnum1] [anum1] [bnum1] · 2
-M [Lnum2] [anum2] [bnum2] (38)
Y [Lnum] [anum] [bnum]
= Y [Lnum1] [anum1] [bnum1] · 2
-Y [Lnum2] [anum2] [bnum2] (39)
If the above processing is performed for all grid points, CMY values of grid points around the grid points in the color gamut can be set.

FIG. 16 is a flowchart illustrating a method for obtaining out-of-gamut CMY values by extrapolation processing.
In FIG. 16, first, as an initial setting, an extrapolation set flag is set to 0 (step S41). In the initial state, data is set only for grid points in the color gamut.

Next, by selecting the L ^* a ^* b ^* number (Lnum, anum, bnum; 0 ≦ Lnum ≦ 7, anum ≦ 13, 0 ≦ bnum ≦ 13), the lattice points of the L ^* a ^* b ^* space are selected. It is selected (step S42), and it is investigated whether data is set at the selected grid point (step S43). If no data is set for the selected grid point, a data setting process is performed.

In the data setting process, a grid point adjacent to the selected grid point and a grid point adjacent to the selected grid point are selected (step S44), and it is checked whether data of the grid point adjacent to the selected grid point and the grid point adjacent to the selected grid point has been set. (Step S45). Here, the L ^* a ^* b ^* number of the adjacent lattice point is (Lnum1, anum1, bnum1), and the L ^* a ^* b ^* number of the adjacent lattice point is (Lnum2, anum2, bum2).

  Then, in any of the 26 directions, when the data of the grid point adjacent to the selected grid point and the grid point adjacent to the selected grid point are set, the grid point adjacent to the selected grid point and the grid point adjacent to the selected grid point The CMY value corresponding to the selected grid point is calculated by performing extrapolation calculation processing using the data (step S46), and the extrapolation set flag is set to 1 (step S47).

On the other hand, if the data of the grid point adjacent to the selected grid point and the data of the grid point adjacent to the selected grid point are not set in all the 26 directions, all the grid points have not been selected (step S48). Returning to S42, the next lattice point in the L ^* a ^* b ^* space is selected.

When the first selection of all grid points is completed and the extrapolation set flag is 1 (step S49), the process returns to step S41, and a new selection is made by performing the second selection of all grid points. The data of the grid point is set further outside the set grid point. When the above process is repeated and the setting of data for all grid points in the L ^* a ^* b ^* space is completed, the extrapolation set flag is determined to be 0 in step S49, and the process ends.

  In this way, only the grid points around the grid points in the printer color gamut can be set using the data of the grid points in the printer color gamut. A CMY value cannot be set only by using a grid point in the color gamut of the printer. For this reason, for grid points far from the grid points in the printer color gamut, first, CMY values are set for grid points around the grid points in the printer color gamut, and the CMY values thus set are further used. Repeat extrapolation outward.

FIG. 17 is a diagram showing an example of the contents of the color conversion table generated by the above processing.
In FIG. 17, assuming that the CMY values that can be reproduced by the printer are in the range of 0 to 255, the data in the range of 0 to 255 is the data in the color gamut, and the range of 0 to 255 in the CMY value. Outside data is data outside the color gamut. Since the data outside the color gamut has a numerical value outside the range of 0 to 255, even if the color conversion between the outside of the color gamut and the inside of the color gamut is performed by interpolation, the data can be accurately converted to the borderline within the color gamut.

  In the above-described embodiment, data is set up to a point that is significantly out of the color gamut. However, data that is greatly out of the color gamut is created by extrapolating a long distance, and thus greatly out of the color gamut. There is a risk that the data that is included contains large errors. For this reason, the degree of deviation from the color gamut of data created by extrapolation may be added as identification data. For example, in the flowchart of FIG. 16, by repeating the process at the branch of “extrapolation set flag = 1” in step S49, data of grid points on the outer side is generated, and data greatly deviating from the color gamut is set. Like to do. Therefore, when grid point data is set, the number of passes of this branch may be added as identification data as it is.

  Also, if the accuracy of the data registered in the color conversion table is low, the accuracy when creating data outside the color gamut will be reduced. The data may be created. For example, when the degree of coincidence with a grid point is described in the color conversion table and data with a low degree of coincidence is registered in the color conversion table, this data is treated as not registered in the color conversion table. As a result, extrapolation using data with low accuracy can be prevented, and generation of out-of-gamut data with low accuracy can be prevented.

Next, a description will be given of a second embodiment of a method for creating a conversion table for converting L ^* a ^* b ^* values into CMY values.
FIG. 18 is a diagram showing a color signal generation process in the color space according to the second embodiment of the color conversion table creation method. In the first embodiment described above, the regularly distributed L ^* a ^* b ^* space is the target of extrapolation processing, but in the second embodiment, the regularly distributed CMY space is extrapolated. Target of processing.

  In FIG. 18, colors regularly distributed as CMY values are printed by a printer. The colors printed by this printer have a regular grid-like distribution in the CMY space and exist only within the color reproduction range of the printer.

Next, the color measurement is performed on the result of printing by this printer. The colors measured by this colorimeter are distributed in the L ^* a ^* b ^* space, and the distribution of L ^* a ^* b ^* values corresponding to the CMY values of the grid points in the CMY space is irregular. That is, correspondence between the L ^* a ^* b ^* values distributed irregularly in CMY values and L ^* a ^* b ^* space which regularly distributed on the CMY space is obtained. The correspondence obtained here is distributed only within the color reproduction range of the printer.

Next, the correspondence between the CMY values and L ^* a ^* b ^* values within the color reproduction range of the printer by extrapolating, CMY values outside the color reproduction range of the printer and L ^* a ^* b ^* values and the Generate a correspondence.

Next, conversion is performed to regularly distribute L ^* a ^* b ^* values that are irregularly distributed in the L ^* a ^* b ^* space. This transformation, L ^* a ^* b ^* performs interpolation of L ^* a ^* b ^* values irregularly distributed in space, L ^* a ^* b ^* values at lattice points in space L ^* a ^* b ^* value This can be done by obtaining the correspondence between CMY values and CMY values. As this interpolation calculation, there is a method in which a plurality of set CMY values are selected so as to surround unset CMY values, and a weight calculation is performed on L ^* a ^* b ^* values corresponding to the CMY values. Then, among the L ^* a ^* b ^* values obtained by interpolation, the one on the lattice point in the L ^* a ^* b ^* space is selected, and the CMY corresponding to the L ^* a ^* b ^* value is selected. Register the value in the color conversion table.

As described above, according to the second embodiment of the color conversion table creation method, the correspondence relationship between the color signals outside the device color reproduction range is obtained by extrapolation from the actual measurement value, and then the L ^* a ^* b ^* space is obtained. By interpolating so that the color signal is distributed at the grid points, it is possible to prevent errors from accumulating when performing extrapolation. It is possible to generate a color conversion table capable of performing color conversion with high accuracy.

Hereinafter, a method for extrapolating the CMY values within the color reproduction range of the printer outside the color reproduction range of the printer will be described.
In this extrapolation process,
(1) Processing for selecting the outermost grid point to be extrapolated,
(2) processing for determining the extrapolation direction;
(3) Processing for selecting inner grid points used in extrapolation calculation,
(4) Process for calculating distance for extrapolation (5) Extrapolation calculation process,
Is done.

Hereinafter, for ease of explanation, it is assumed that a total of 125 colors are printed and measured for the following five levels of C value, Y value, and M value.
C value = (int) ((Cnum−1) · 255 + 2) / 4 (40)
M value = (int) ((Mnum−1) · 255 + 2) / 4 (41)
Y value = (int) ((Ynum-1) .255 + 2) / 4 (42)
However, Cnum, Mnum, and Ynum are integer values of 1 to 5, respectively, and (int) indicates truncation after the decimal point.

According to the equations (40) to (42), the CMY values can be labeled with CMY numbers (Cnum, Mnum, Ynum), and the L ^* a ^* b ^* values that are the colorimetric values are also L ^* a ^*. It can be labeled with b ^* numbers (Lnum, anum, bnum). Here, the correspondence obtained by colorimetry is 5 × 5 × 5 data, but by creating data outside the color gamut around the color gamut obtained by colorimetry, 7 × 7 × 7 Data shall be generated. In this case, the storage format of the L ^* a ^* b ^* values is a three-dimensional array (L [7] [7] [7], a [7] [7] with CMY numbers (Cnum, Mnum, Ynum) as arguments. ] [7], b [7] [7] [7]). Data obtained by colorimetry is stored in an array having argument values 1 to 5, and data generated by extrapolation is stored in portions where the argument values are 0 and 6.
(1) In the process of selecting the outermost grid point to be extrapolated, the outermost grid point is selected from the data obtained by colorimetry arranged in a 5 × 5 × 5 grid pattern in the CMY space To do.

Specifically, as shown below, there are three types of selection methods of a cubic surface, a ridge line, and a vertex, and there are points that are selected redundantly on the surface and the ridge line. Note that * is an arbitrary value from 1 to 5.
(I) For a cube surface, select all outer grid points and represent them by array arguments:
(1, *, *), (5, *, *), (*, 1, *),
(*, 5, *), (*, 1, *), (*, *, 5)
It becomes.
(Ii) For the ridgeline of a cube, if all the lattice points on the ridgeline are selected and represented by an array argument,
(1,1, *), (1,5, *), (5,1, *),
(5, 5, *), (*, 1, 1), (*, 1, 5),
(*, 5, 1), (*, 5, 5), (1, *, 1),
(1, *, 5), (5, *, 1), (5, *, 5),
It becomes.
(Iii) For the vertices of the cube, if all the lattice points on the vertices are selected and expressed by array arguments,
(1,1,1), (1,1,5), (1,5,1), (5,1,1),
(5, 5, 1), (5, 1, 5), (1, 5, 5), (5, 5, 5),
It becomes.
(2) In the process of determining the extrapolation direction, the extrapolation direction is determined for each of the cube surface, ridge line, and vertex.

FIG. 19 is a diagram illustrating the direction of extrapolation.
In FIG. 19, the measured CMY values are distributed at 5 × 5 × 5 cube lattice points, and the type of extrapolation direction is the direction perpendicular to the surface (surface extrapolation) for this cube, Directions that are perpendicular to the ridgeline and equiangular to the two faces (use ridgeline data (ridgeline extrapolation)), and equiangular to the three faces of the cube (use vertex data (vertex extrapolation)) There are three types.

If extrapolation in these directions is performed on the entire 5 × 5 × 5 grid, for example, data distributed in a 5 × 5 × 5 grid becomes data distributed in a 7 × 7 × 7 grid.
(3) In the process of selecting the inner grid point used in the extrapolation calculation, basically, the relationship between the outermost data of the color reproduction range and the data inside the one is used to calculate the innermost grid point. Create data outside the outside. For example, in FIG. 19, in vertex extrapolation, data of a point P1 outside the vertex P2 is generated by using data of the cube vertex P2 and the point P3 inside it. In the surface extrapolation, the data of the point P4 outside the point P5 is generated by using the data of the point P5 on the surface of the cube and the point P6 inside it. In ridge extrapolation, data of a point P7 outside the point P8 is generated by using data of the point P8 on the ridge line of the cube and the point P9 inside the cube.

  However, in exceptional cases, the data inside rather than one inside may be used. This is to ensure accuracy when color collapse occurs (when the color difference between the created color charts (patches) is small).

  Specifically, one inner lattice point is selected according to each extrapolation direction of FIG. Here, the array argument of the outermost lattice is (C0, M0, Y0), and the array argument of the inner lattice is (C1, M1, Y1).

  If the following conditions are not satisfied for the selected inner grid point, a further inner grid point is selected. Then, by repeating the inward selection process, a grid point that satisfies the condition is selected, and the array arguments (C1, M1, Y1) are updated.

  Note that the limit of the inner selection range is half the number of colors (rounded down) for each color of CMY at the grid points. For example, since the number of colors made up of 5 × 5 × 5 grid points is 5, it is assumed that the number is 2 inside. If there is no grid point that satisfies the condition in this range, the innermost one up to the limit of the selection range is selected.

FIG. 20 is a diagram for explaining a method of selecting inner grid points.
In FIG. 20, grid points P1 to P9 in CMY space correspond to points Q1 to Q9 in L ^* a ^* b ^* space, and grid points P1 to P3 are on the boundary of the color reproduction range. And In this case, the lattice point P4 is selected as a point inside the lattice point P1, and the distance D1 between the point Q1 and the point Q4 in the L ^* a ^* b ^* space corresponding to the lattice points P1 and P4, respectively, The angle θ1 formed by the straight line connecting the point Q1 and the point Q4 and the straight line connecting the point Q1 and the point R (50, 0, 0) is examined. An angle formed by a straight line connecting point Q1 and point R (50, 0, 0) and a straight line connecting point Q1 and point Q4 with distance D1 between point Q1 and point Q4 being equal to or greater than threshold value TH1 When θ1 is less than or equal to threshold value TH2, extrapolation is performed by using point Q1 and point Q4.

Further, the grid point P5 is selected as a point inside the grid point P2, and the distance D2 between the point Q2 and the point Q5 in the L ^* a ^* b ^* space corresponding to the grid points P2 and P5, respectively, and the point The angle θ2 formed by the straight line connecting Q2 and point Q5 and the straight line connecting point Q2 and point R (50, 0, 0) is examined. Here, when the angle θ2 formed by the straight line connecting the point Q2 and the point Q5 and the straight line connecting the point Q2 and the point R (50, 0, 0) is large and exceeds the threshold value TH2, the lattice point P2 A grid point P8 further inside the grid point P5 is selected as a point inside the grid point P5.

The distance between the point Q2 and the point Q8 in the L ^* a ^* b ^* space corresponding to the lattice points P2 and P8, the straight line connecting the point Q2 and the point Q8, the point Q2 and the point R (50 , 0, 0) and the angle formed by the straight line connecting them is examined. As a result, the distance between the point Q2 and the point Q8 is equal to or greater than the threshold value TH1, and the straight line connecting the point Q2 and the point Q8 and the straight line connecting the point Q2 and the point R (50, 0, 0). When the formed angle is equal to or smaller than the threshold value TH2, extrapolation is performed by using the point Q2 and the point Q8.

Further, the grid point P6 is selected as a point inside the grid point P3, and the distance D3 between the point Q3 and the point Q6 in the L ^* a ^* b ^* space corresponding to the grid points P3 and P6, respectively, An angle θ3 formed by a straight line connecting Q3 and point Q6 and a straight line connecting point Q3 and point R (50, 0, 0) is examined. Here, when the distance D3 between the point Q3 and the point Q6 is small and does not exceed the threshold value TH1, a lattice point P9 further inside the lattice point P6 is selected as a point inside the lattice point P3. Then, the distance between the points Q3 and Q9 in the L ^* a ^* b ^* space corresponding to the lattice points P3 and P9, the straight line connecting the points Q3 and Q9, the point Q3 and the point R (50 , 0, 0) and the angle formed by the straight line connecting them is examined. As a result, the distance between the point Q3 and the point Q9 is equal to or greater than the threshold value TH1, and the straight line connecting the point Q3 and the point Q9 and the straight line connecting the point Q3 and the point R (50, 0, 0). When the angle formed is equal to or less than the threshold value TH2, extrapolation is performed by using the point Q3 and the point Q9.

Note that the point R (50, 0, 0) in the L ^* a ^* b ^* space is an achromatic and gray point, and the direction toward the selected point and the direction toward the point R (50, 0, 0). In general, the points from the outside to the inside are gathered in gray and gray points, so the direction toward the selected point is from the direction toward the point R (50, 0, 0). This is because the error is considered to be large if it is off.

  The reason why the interpolation calculation is not performed using the two points having a small distance is that if the extrapolation is performed using the two points having a small distance, it is necessary to increase the extrapolated distance, and the error increases. Because.

Hereinafter, a method for selecting the inner grid point will be described using mathematical expressions.
First, the following dist, dist ′, and angle are calculated in the L ^* a ^* b ^* space. dist and dist ′ are distances of L ^* a ^* b ^* values of the two data, and angle is a vector connecting the point (L0, a0, b0) and the point (50, 0, 0), This is the inner product of the vector connecting the point (L0, a0, b0) and the point (L1, a1, b1), and the value decreases as the angle formed by these two vectors increases. Here, (L0, a0, b0) and (L1, a1, b1) are
L0 = L [C0] [M0] [Y0] (43)
a0 = a [C0] [M0] [Y0] (44)
b0 = b [C0] [M0] [Y0] (45)
L1 = L [C1] [M1] [Y1] (46)
a1 = a [C1] [M1] [Y1] (47)
b1 = b [C1] [M1] [Y1] (48)
Is represented by Also, dist, dist1, and angle are
dist = sqrt ((L0−L1) ² + (a0−a1) ²
+ (B0−b1) ² ) (49)
dist '= sqrt ((L0-50) ² + (a0) ² + (b0) ² )
... (50)
if (dist == 0.0, or dist ′ == 0.0)
angle = 0.0 (51)
else
angle = ((L0−L1) · (L0−50) + (a0−a1) · a0
+ (B0−b1) · b0) / (dist · dist1)
... (52)
Is represented by

Next, check whether the following conditions are met or not.
dist> TH1 and angle> TH2 (53)
Through the above process,
1. CMY number of inner grid point (C1, M1, Y1)
2. 2. L ^* a ^* b ^* values corresponding to two grid points in the CMY space A value w2 (basically 1) indicating how many CMY numbers (C1, M1, Y1) are inside of the CMY numbers (C0, M0, Y0)
4). Distance dist between two points in L ^* a ^* b ^* space
Can be requested.
(4) In the process of calculating the extrapolated distance, it is determined how many times the distance dist between the two points in the L ^* a ^* b ^* space is to be generated. If extrapolation is performed using the distance dist between the two points as it is, the extrapolation points are concentrated in the color reproduction range boundary portion of the L ^* a ^* b ^* space when the distance dist between the two points is short. This is because there is a risk that data that completely covers the reproduction range is not registered in the color conversion table.

FIG. 21 is a diagram for explaining a method of calculating a multiple of the distance dist between two points to be extrapolated.
In FIG. 21, lattice points P1 to P10 in the CMY space correspond to points Q1 to Q10 in the L ^* a ^* b ^* space, and the lattice points P6 to P10 in the CMY space are on the boundary of the color reproduction range. It shall be. Further, the lattice point P1 is selected as the lattice point inside the lattice point P6, the lattice point P2 is selected as the lattice point inside the lattice point P7, and the lattice point P3 is selected as the lattice point inside the lattice point P8. Assume that the lattice point P4 is selected as the lattice point inside the lattice point P9, and the lattice point P5 is selected as the lattice point inside the lattice point P10. Then, by using equation (49), corresponding to the grid points ^{P1, P6 L * a * b} * distance between the point Q1 and point Q6 space dist1, corresponding to the grid points P2, P7 L ^* a ^* b ^* The distance dist2 between the point Q2 and the point Q7 in the space, the distance dist3 between the point Q3 and the point Q8 in the L ^* a ^* b ^* space corresponding to the lattice points P3 and P8, and the lattice points P4 and P9 the distance between the distance Dist4, L ^* a ^* b ^* point of space Q5 and the point Q10 corresponding to the grid points P5, P10 between the L ^* a ^* b ^* point of space Q4 and the point Q9 corresponding to dist5 Are obtained.

Here, the distance of extrapolation is basically equal to each dist, but when the dist is short, a distance that is an integral multiple of dist is extrapolated. That is, a multiple w1 indicating how many times the distance of dist is extrapolated is
w1 = (int) (g · gex / dist) +1 (54)
It can ask for. Here, g is a lattice interval in the L ^* a ^* b ^* space, and gex is a predetermined constant.

  For example, if the distance dist1 between the point Q1 and the point Q6 is substantially equal to the value of g · gex, the multiple w1 is 1, so that the point Q11 that is separated from the point Q6 by the distance dist1 is outside. Use as an insertion point.

  If the distance dist2 between the point Q2 and the point Q7 is equal to ½ of the value of g · gex, the multiple w1 is 2, so that the distance from the point Q7 is twice the distance dist2. A point Q12 that is separated outward is set as an extrapolation point.

  If the distance dist3 between the point Q3 and the point Q8 is equal to 1/4 of the value of g · gex, the multiple w1 is 4, so that the distance from the point Q8 is four times the distance dist3. A point Q13 that is separated outward is set as an extrapolation point.

  Further, if the distance dist4 between the point Q4 and the point Q9 is equal to 1/5 of the value of g · gex, the multiple w1 is 5, so that the distance from the point Q9 is 5 times the distance dist4. A point Q14 that is separated outward is set as an extrapolation point.

  Further, if the distance dist5 between the point Q5 and the point Q10 is equal to 1/7 of the value of g · gex, the multiple w1 is 7, so that the distance from the point Q10 is 7 times the distance dist5. A point Q15 that is separated outward is set as an extrapolation point.

Note that the point in the CMY space corresponding to the point Q11 in the L ^* a ^* b ^* space is made to correspond to the lattice point P11 outside by the lattice interval f in the CMY space because the point Q11 is located outside by the distance dist1.

Further, the point in the CMY space corresponding to the point Q12 in the L ^* a ^* b ^* space is outside the point Q12 by a distance twice as long as the distance dist2, so that the point is outside the lattice spacing f in the CMY space by twice. Corresponding to the grid point P11.

Also, the point in the CMY space corresponding to the point Q13 in the L ^* a ^* b ^* space is located outside the point Q13 by an amount that is four times the distance dist3, and thus is four times the lattice spacing f in the CMY space. Corresponding to the grid point P13.

In addition, the point in the CMY space corresponding to the point Q14 in the L ^* a ^* b ^* space is located outside the point Q14 by an amount 5 times the distance dist4, and thus is 5 times the lattice spacing f in the CMY space. Corresponding to the grid point P14.

Also, the point in the CMY space corresponding to the point Q15 in the L ^* a ^* b ^* space is outside the point Q15 by 7 times the distance dist5, so that it is outside by 7 times the lattice spacing f in the CMY space. Corresponding to the grid point P15.
(5) In the extrapolation calculation, by using the grid point at the boundary of the color reproduction range in the CMY space and the grid point inside the CMY space, the outside of the distance dist between the two points in the L ^* a ^* b ^* space by w1 times. The L ^* a ^* b ^* value corresponding to the point is obtained.

  Hereinafter, in order to simplify the description, the CMY number indicating the position of the grid point that is only one outside the grid point indicated by the CMY number (C0, M0, Y0) is (C2, M2, Y2), and the outer side is w2. The CMY numbers indicating the positions of the grid points are (C3, M3, Y3). For example, the CMY number (C0, M0, Y0) is (1, 3, 3), the CMY number (C1, M1, Y1) is (3, 3, 3) (in this case, w2 = 2), and w1 = In the case of 3, the CMY number (C2, M2, Y2) = (0, 3, 3) and the CMY number (C3, M3, Y3) = (− 2, 3, 3).

Here, the CMY number (C2, M2, Y2) and the CMY number (C3, M3, Y3) are
C2 = C0− (C1−C0) / w2 (55)
M2 = M0− (M1−M0) / w2 (56)
Y2 = Y0- (Y1-Y0) / w2 (57)
C3 = C0− (C2−C0) · w1 (58)
M3 = M0− (M2−M0) · w1 (59)
Y3 = Y0− (Y2−Y0) · w1 (60)
It can ask for.

Also,
L0 = L [C0] [M0] [Y0] (61)
a0 = a [C0] [M0] [Y0] (62)
b0 = b [C0] [M0] [Y0] (63)
L1 = L [C1] [M1] [Y1] (64)
a1 = a [C1] [M1] [Y1] (65)
b1 = b [C1] [M1] [Y1] (66)
Then, the L ^* a ^* b ^* value corresponding to the CMY number (C2, M2, Y2) is
L [C2] [M2] [Y2] = L0−w1 · (L1−L0) (67)
a [C2] [M2] [Y2] = a0-w1. (a1-a0) (68)
b [C2] [M2] [Y2] = b0−w1 · (b1−b0) (69)
It can ask for.

In addition, the position of the grid point outside by w2 is
C3 [C2] [M2] [Y2] = C3 (70)
M3 [C2] [M2] [Y2] = M3 (71)
Y3 [C2] [M2] [Y2] = Y3 (72)
It becomes.

As described above, generation of data outside the color gamut by extrapolation is completed, but the L ^* a ^* b ^* values in the L ^* a ^* b ^* space are irregularly distributed. the distributed L ^* a ^* b ^* values are converted into L ^* a ^* b ^* values are distributed in a grid shape. This process is basically the same as the process shown in FIG. That is, by using the correspondence between the CMY numbers distributed in a lattice pattern and the L ^* a ^* b ^* values distributed irregularly, the L ^* a ^* b ^* values are interpolated to obtain the CMY values and Increase the correspondence with L ^* a ^* b ^* values. For example, a unit cube generated by eight grid points in the CMY space is divided into 64 × 64 × 64, and L ^* a ^* b ^* values corresponding to the CMY values of the division points are divided into eight pieces surrounding the division points. It is obtained by weight calculation of L ^* a ^* b ^* values corresponding to lattice points.

Next, a lattice point having an L ^* a ^* b ^* value close to the L ^* a ^* b ^* value obtained by the weight calculation is detected from the L ^* a ^* b ^* space, and the L ^* a ^* corresponding to this lattice point is detected ^. b ^* Calculate the number.
Then, calculates the square of the distance between the L ^* a ^* b ^* values of grid points that are close to the L ^* a ^* b ^* value obtained by the weight calculation in the L ^* a ^* b ^* values.

Next, by sequentially comparing the squares of the distances between the L ^* a ^* b ^* values obtained by the weight calculation and the L ^* a ^* b ^* values of the lattice points for all the division points of the unit cube, A division point having an L ^* a ^* b ^* value closest to the L ^* a ^* b ^* value is detected. Then, the CMY value of the dividing point is registered as a value corresponding to the L ^* a ^* b ^* value of the grid point.

Hereinafter, it demonstrates more concretely using numerical formula.
Since the data to be created are CMY values corresponding to L ^* a ^* b ^* numbers (Lnum, annum, bnum) of lattice points in the L ^* a ^* b ^* space, the actual created data format is
C [Lnum] [anum] [bnum],
M [Lnum] [anum] [bnum],
Y [Lnum] [anum] [bnum],
This is a three-dimensional array.

  First, assuming that a 5 × 5 × 5 lattice point in the CMY space becomes a 7 × 7 × 7 lattice point by extrapolation, 0 ≦ Cnum ≦ 6, 0 ≦ Mnum ≦ 6, 0 ≦ Ynum. For all combinations of integer values ≦ 6, grid points with CMY numbers (Cnum, Mnum, Ynum) are sequentially selected.

  Next, for the integer values (Cw, Mw, Yw) in the range of 0 ≦ Cw ≦ W, 0 ≦ Mw ≦ W, 0 ≦ Yw ≦ W, the colors (Lnew, anew, bnew) by interpolation are expressed by the following equations. create.

Lnew = (L [Cnum] [Mnum] [Ynum] · V (0,0,0)
+ L [Cnum + 1] [Mnum] [Ynum] · V (1, 0, 0)
+ L [Cnum] [Mnum + 1] [Ynum] · V (0,1,0)
+ L [Cnum] [Mnum] [Ynum + 1] · V (0, 0, 1)
+ L [Cnum] [Mnum + 1] [Ynum + 1] · V (0,1,1)
+ L [Cnum + 1] [Mnum] [Ynum + 1] · V (1, 0, 1)
+ L [Cnum + 1] [Mnum + 1] [Ynum] · V (1,1,0)
+ L [Cnum + 1] [Mnum + 1] [Ynum + 1] · V (1,1,1)
) / (W / W / W) (73)
anew = (a [Cnum] [Mnum] [Ynum] · V (0,0,0)
+ A [Cnum + 1] [Mnum] [Ynum] · V (1, 0, 0)
+ A [Cnum] [Mnum + 1] [Ynum] · V (0,1,0)
+ A [Cnum] [Mnum] [Ynum + 1] · V (0, 0, 1)
+ A [Cnum] [Mnum + 1] [Ynum + 1] · V (0,1,1)
+ A [Cnum + 1] [Mnum] [Ynum + 1] · V (1, 0, 1)
+ A [Cnum + 1] [Mnum + 1] [Ynum] · V (1,1,0)
+ A [Cnum + 1] [Mnum + 1] [Ynum + 1] · V (1,1,1)
) / (W / W / W) (74)
bnew = (b [Cnum] [Mnum] [Ynum] · V (0,0,0)
+ B [Cnum + 1] [Mnum] [Ynum] · V (1, 0, 0)
+ B [Cnum] [Mnum + 1] [Ynum] · V (0,1,0)
+ B [Cnum] [Mnum] [Ynum + 1] · V (0, 0, 1)
+ B [Cnum] [Mnum + 1] [Ynum + 1] · V (0,1,1)
+ B [Cnum + 1] [Mnum] [Ynum + 1] · V (1, 0, 1)
+ B [Cnum + 1] [Mnum + 1] [Ynum] · V (1,1,0)
+ B [Cnum + 1] [Mnum + 1] [Ynum + 1] · V (1,1,1)
) / (W / W / W) (75)
here,
V (0,0,0) = (W-Cw). (W-Mw). (W-Yw)
... (76)
V (1, 0, 0) = Cw · (W−Mw) · (W−Yw) (77)
V (0,1,0) = (W−Cw) · Mw · (W−Yw) (78)
V (0,0,1) = (W−Cw) · (W−Mw) · Yw (79)
V (0,1,1) = (W−Cw) · Mw · Yw (80)
V (1, 0, 1) = Cw · (W−Mw) · Yw (81)
V (1,1,0) = Cw · Mw · (W−Yw) (82)
V (1,1,1) = Cw · Mw · Yw (83)
It is. However, W is a preset value, and is 64 here. Note that V (0,0,0), V (1,0,0), V (0,1,0) V (0,0,1), V (0,1,1), V (1, 0,1), V (1,1,0), V (1,1,1) pass through a cube with one side length W passing through the point of (Cw, Mw, Yw), and to each CMY axis. It represents the volume of eight rectangular parallelepipeds when divided on a vertical plane. Expressions (73) to (75) are obtained by multiplying L ^* a ^* b ^* values respectively corresponding to the eight vertices of the unit cube in the CMY space by the weights of expressions (76) to (83). It corresponds to.

Next, the grid number near the color (Lnew, anew, bnew) by interpolation is selected by the following formula.
G is the lattice spacing in the L ^* a ^* b ^* space.

anew + = 128
bnew + = 128
Lnum = (Lnew / G> 0.0?
(Int) (Lnew / G + 0.5):
(Int) (Lnew / G-0.5)), (85)
anum = (anew / G> 0.0?
(Int) (anew / G + 0.5):
(Int) (anew / G-0.5)), (86)
bnum = (bnew / G> 0.0?)
(Int) (bnew / G + 0.5):
(Int) (bnew / G-0.5)), (87)
When all of Lnum, anum, and bnum satisfy the following conditions, the distance from the lattice point is calculated.

0 ≦ Lnum ≦ (int) (100 / G) +2 (88)
0 ≦ anum ≦ (int) (256 / G) +1 (89)
0 ≦ bnum ≦ (int) (256 / G) +1 (90)
Next, the square of the distance between the interpolated color (Lnew, anew, bnew) and the selected grid point is calculated by the following equation.

diff = (Lnum−Lnew / G) ²
+ (Anum-anew / G) ²
+ (Bnum-bnew / G) ² (91)
Only when diff satisfies the following conditions, data setting processing according to the distance is performed.

diff ≦ D (92)
However, D is a predetermined value.
Next, there is no data set at the lattice point of (Lnum, annum, bnum) or diff is already set at the lattice point of (Lnum, annum, bnum) diff [Lnum] [anum] If it is smaller than [bnum], CMY values corresponding to the lattice points of (Lnum, annum, bnum) are calculated, and these data are set.

  Here, when Cnum, Mnum, and Ynum are not all 0 and 6 (within the color gamut), a cube is formed with eight vertices surrounding the division points in the CMY space. Can be sought.

C [Lnum] [anum] [bnum]
= (Cnum-1) · G + Cw · G / W (93)
M [Lnum] [anum] [bnum]
= (Mnum-1) · G + Mw · G / W (94)
Y [Lnum] [anum] [bnum]
= (Ynum-1) · G + Yw · G / W (95)
Also, diff is set as follows.

diff [Lnum] [anum] [bnum] = diff
On the other hand, if any one of C, M, and Y has 0 or 6 (out of the color gamut), a cube is not always formed by eight vertices surrounding the division point of the CMY space. Is obtained by interpolation.

That is, the CMY grid numbers of 8 colors used for interpolation are calculated by the following formula.
Cn [0] = Cnum (96)
Mn [0] = Mnum (97)
Yn [0] = Ynum (98)
Cn [1] = Cn [0] +1 (99)
Mn [1] = Mn [0] (100)
Yn [1] = Yn [0] (101)
Cn [2] = Cn [0] (102)
Mn [2] = Mn [0] +1 (103)
Yn [2] = Yn [0] (104)
Cn [3] = Cn [0] (105)
Mn [3] = Mn [0] (106)
Yn [3] = Yn [0] +1 (107)
Cn [4] = Cn [0] (108)
Mn [4] = Mn [0] +1 (109)
Yn [4] = Yn [0] +1 (110)
Cn [5] = Cn [0] +1 (111)
Mn [5] = Mn [0] (112)
Yn [5] = Yn [0] +1 (113)
Cn [6] = Cn [0] +1 (114)
Mn [6] = Mn [0] +1 (115)
Yn [6] = Yn [0] (116)
Cn [7] = Cn [0] +1 (117)
Mn [7] = Mn [0] +1 (118)
Yn [7] = Yn [0] +1
Next, when each lattice point (Cn [x], Mn [x], Yn [x]) obtained by the equations (96) to (118) is a lattice point created by extrapolation, only w2 is obtained. Replace with the position of the outer grid. (The following processing is executed from x = 0 to x = 7)
if (Cn [x] = 0, or Cn [x] = 6, or Mn [x] = 0,
Or Mn [x] = 6, or Yn [x] = 0, or Yn [x] = 6)
Cn [x] = C3 [Cn [x]] (119)
Mn [x] = M3 [Mn [x]] (120)
Yn [x] = Y3 [Yn [x]] (121)
Next, CMY values are created by interpolation as shown in the following formulas (122) to (124) by using the weighting coefficients of formulas (76) to (83).

C [Lnum] [anum] [bnum]
= (Cn [0] · V (0,0,0) + Cn [1] · V (1,0,0)
+ Cn [2] · V (0,1,0) + Cn [3] · V (0,0,1)
+ Cn [4] · V (0,1,1) + Cn [5] · V (1,0,1)
+ Cn [6] · V (1,1,0) + Cn [7] · V (1,1,1)
) / (W · W · W) (122)
M [Lnum] [anum] [bnum]
= (Mn [0] · V (0,0,0) + Mn [1] · V (1,0,0)
+ Mn [2] · V (0,1,0) + Mn [3] · V (0,0,1)
+ Mn [4] · V (0,1,1) + Mn [5] · V (1,0,1)
+ Mn [6] · V (1,1,0) + Mn [7] · V (1,1,1)
) / (W / W / W) (123)
Y [Lnum] [anum] [bnum]
= (Yn [0] · V (0,0,0) + Yn [1] · V (1,0,0)
+ Yn [2] · V (0,1,0) + Yn [3] · V (0,0,1)
+ Yn [4] · V (0,1,1) + Yn [5] · V (1,0,1)
+ Yn [6] · V (1,1,0) + Yn [7] · V (1,1,1)
) / (W / W / W) (124)
Through the above processing, it is possible to generate a color conversion table having data outside the color gamut and having L ^* a ^* b ^* values distributed in a grid pattern. Since the CMY value of the data outside the color gamut has a value outside the range of 0 to 255, even when the color conversion is performed by interpolation between the color outside the color gamut and the color gamut, the color conversion is accurately performed to the very minimum within the color gamut. It becomes possible.

  In order to generate data for points that deviate significantly from the color gamut, it is necessary to extrapolate a long distance. However, since the error increases as the extrapolation distance increases, the data deviates greatly from the color gamut. However, by not generating data, it may be possible to prevent generation of data that may increase the error. However, it is possible to set data at all grid points by extrapolating a long distance.

FIG. 22 is a flowchart showing the process according to the second embodiment of the color conversion table creation method.
In FIG. 22, first, the correspondence between the color signals in the color gamut distributed at the grid points in the CMY space and the color signals in the L ^* a ^* b ^* space is obtained (step S51), and distributed to the grid points in the CMY space. Out of the color signals in the color gamut, two outer color signals are selected (step S52).

Next, the extrapolated distance of the L ^* a ^* b ^* space is calculated based on the two selected color signals (step S53), and the color signal outside the color gamut is placed outside by the distance calculated in step S53. Generate (step S54). Next, by interpolating between the color signals in the color gamut distributed at the grid points in the CMY space and the color signals outside the color gamut, a color signal distributed at the grid points in the L ^* a ^* b ^* space is generated (step S55). ).

FIG. 23 is a flowchart showing color signal generation processing outside the color gamut.
In FIG. 23, first, outer grid points are sequentially selected from the grid points in the color gamut of the CMY space (step S61), and a grid point inside one of the selected outer grid points is selected (step S62). ).

  Next, it is checked whether or not the condition of the inner grid point in the equation (53) is satisfied (step S63), and when (int) (CMYnum / 2) inner grid points are not selected to the inner side (step S63) In step S64), an inner grid point is selected (step S65). Here, CMYnum is the colorimetric number of each CMY component, and is 5 in the example of FIG. 19, for example.

If the condition of the inner grid point in the equation (53) is satisfied, L ^* is determined based on the color signal of the outer grid point belonging to the color gamut of the CMY space and the color signal of the selected inner grid point ^. The extrapolated distance of the a ^* b ^* space is calculated (step S66).

Next, a color signal out of the color gamut is generated outside the distance calculated in step S66 by the equations (67) to (69) (step S67), and L ^* is calculated by the equations (70) to (72) ^. The positions of grid points in the CMY space corresponding to the color signals generated by extrapolation are stored in the a ^* b ^* space.

The above processing is performed for all outer grid points belonging to the color gamut of the CMY space, and when selection of all outer grid points is completed, the color signal generation processing outside the color gamut ends.
FIG. 24 is a flowchart showing color signal generation processing of lattice points in the L ^* a ^* b ^* space.

  In FIG. 24, first, by specifying CMY numbers (Cnum, Mnum, Ynum), lattice points in the CMY space are sequentially selected (step S71), and division points (Cw, Mw, Yw) in the CMY space are sequentially selected. (Step S72).

Next, the unit cube including the dividing points (Cw, Mw, Yw) passes through the dividing points (Cw, Mw, Yw) and is divided on a plane perpendicular to each axis of CMY. Then, by using Equations (76) to (83), the volumes of the eight rectangular parallelepipeds obtained at this time are obtained as weighting factors. Thereafter, a new color signal is obtained by using the equations (73) to (75) for the L ^* a ^* b ^* values corresponding to the eight vertices of the unit cube including the dividing points (Cw, Mw, Yw). (Lnew, anew, bnew) is generated (step S73).

Next, the L ^* a ^* b ^* numbers (Lnum, anum, bnum) of the lattice points near the new color signal (Lnew, anew, bnew) are calculated by the equations (85) to (87) (steps). S74) When this L ^* a ^* b ^* number (Lnum, anum, bnum) is within the predetermined range of the equations (88) to (90) (step S75), a new color is obtained by the equation (91). A distance diff between the signal (Lnew, anew, bnew) and the lattice point whose L ^* a ^* b ^* number is (Lnum, annum, bnum) is calculated (step S76).

If the distance diff is within the predetermined range (step S77), it is checked whether or not the distance diff of the lattice point whose L ^* a ^* b ^* number is (Lnum, annum, bnum) is already set (step S78). . If the distance diff of the lattice point with L ^* a ^* b ^* number (Lnum, anum, bnum) is not set, the lattice point with L ^* a ^* b ^* number (Lnum, anum, bnum) is set. If the data is updated (step S80) and the distance diff of the grid point whose L ^* a ^* b ^* number is (Lnum, annum, bnum) is set, the calculated distance diff is larger than the already set value. Is also smaller (step S79).

If the calculated distance diff is smaller than the already set value, the data of the lattice point whose L ^* a ^* b ^* number is (Lnum, annum, bnum) is updated, and the calculated distance diff is already set. If the value is equal to or greater than the specified value, the data of the lattice point whose L ^* a ^* b ^* number is (Lnum, annum, bnum) is left as it is.

The above processing is repeated for all the dividing points in the selected unit cube (step S81), and when the processing is completed for all the dividing points in the selected unit cube, the next CMY number (Cnum, Mnum, Ynum) is obtained. ) Is selected to select the next unit cube, and when all the CMY numbers (Cnum, Mnum, Ynum) are selected, the color signal generation processing of the lattice points in the L ^* a ^* b ^* space is terminated ( Step S82).

Next, a color conversion method using a color conversion table in which color signals outside the color reproduction range are registered will be described. This color conversion table stores out-of-gamut CMY values for L ^* a ^* b ^* grid points. Then, eight points around the L ^* a ^* b ^* values to be converted (values registered in the color conversion table) are selected, and the CMY values for the selected L ^* a ^* b ^* values are interpolated and converted. Obtain the resulting CMY value.

For this reason, if this color conversion table is used, it is possible to accurately obtain the color signal near the boundary of the color reproduction range by interpolation, but the value obtained by interpolation may be outside the color reproduction range. is there. At this time, since a color outside the color reproduction range cannot be printed by the printer, the printer cannot determine which color should be printed with respect to the input L ^* a ^* b ^* value. Therefore, it is determined whether the value obtained by interpolation belongs to the color gamut or the color gamut. If the value belongs to the color gamut, the value outside the color gamut is replaced with the value in the color gamut and output to the printer. To do.

FIG. 25 is a diagram showing a range on a color space of points registered in the color conversion table.
In FIG. 25, the CMY values registered in the color conversion table include a point in the color gamut where the color can be reproduced and a point outside the color gamut where the color cannot be reproduced. Here, a point outside the large color gamut that is further outside the point outside the color gamut is not registered in the color conversion table.

FIG. 26 is a diagram illustrating a method for converting a point outside the color gamut generated by interpolation into a point within the color gamut in a two-dimensional color space.
In FIG. 26, points (black circles) belonging to the color gamut and outside the color gamut are registered in the color conversion table, and points (white circles) belonging to the large color gamut are not registered in the color conversion table. As L ^* a ^* b ^* values to be converted, the L ^* a ^* b ^* value of the point P is assumed to have been entered, L ^* a ^* b ^* value of the point P is to belong to either color gamut within the color gamut Find out if it belongs. In order to investigate this, points P1 to P4 surrounding the point P are extracted. This extraction can be easily performed because the L ^* a ^* b ^* values registered in the color conversion table are distributed in a grid pattern.

When the points P1 to P4 are extracted, it is determined whether the points P1 to P4 belong to the color gamut or the color gamut. Here, the L ^* a ^* b ^* space, and the range is randomly distributed in the color gamut, is so difficult to perform determination whether the color gamut or color range in the L ^* a ^* b ^* space, the color conversion table , The color signals of the points P1 to P4 are converted into the color signals of the points Q1 to Q4 in the CMY space, and then it is determined whether the points P1 to P4 belong to the color gamut or the color gamut. In the CMY space, since the range in the color gamut is distributed in a cubic range, it can be easily determined whether the color gamut is out of the color gamut.

  By this conversion, it is determined that the points P1 and P2 belong to the color gamut and the points P3 and P4 belong to the color gamut. When the points P1 and P2 belong to the color gamut and the points P3 and P4 belong to the outside of the color gamut, the determination result of the points P1 to P4 is used to determine whether the point P belongs to the color gamut or outside the color gamut. It is difficult.

For this reason, a color signal on the CMY space corresponding to the color signal on the L ^* a ^* b ^* space is calculated by performing an interpolation operation. As a result, if the point Q is obtained as a point corresponding to the point P, the range in the color gamut is distributed in the range of the cube in the CMY space. It can be easily determined whether it belongs to the outside of the area.

  Here, if it is determined that the point Q is out of the color gamut, the point Q corresponds to a color outside the color reproduction range of the printer. Therefore, the point Q color signal is simply output to the printer. Printing with a printer becomes impossible. For this reason, the color signal at the point Q is replaced with a color signal in the color gamut before being output to the printer.

In this color signal replacement, a point S in the CMY space corresponding to the point R of the same luminance and achromatic color as the point P is obtained, and among the color signals directed from the point Q to the point S, the point Q ′ on the color gamut boundary. Replace with the color signal.
FIG. 27 is a diagram showing, on a two-dimensional color space, a method for converting a point whose grid point used for interpolation is significantly outside the color gamut into a point within the color gamut.

In FIG. 27, points (black circles) belonging to the color gamut and outside the color gamut are registered in the color conversion table, and points (white circles) belonging to the large color gamut are not registered in the color conversion table. As L ^* a ^* b ^* values to be converted, the L ^* a ^* b ^* value of the point P is assumed to have been entered, L ^* a ^* b ^* value of the point P is to belong to either color gamut within the color gamut Find out if it belongs. In order to investigate this, points P1 to P4 surrounding the point P are extracted. This extraction can be easily performed because the L ^* a ^* b ^* values registered in the color conversion table are distributed in a grid pattern.

When the points P1 to P4 are extracted, it is determined whether the points P1 to P4 belong to the color gamut or the color gamut. Here, the L ^* a ^* b ^* space, and the range is randomly distributed in the color gamut, is so difficult to perform determination whether the color gamut or color range in the L ^* a ^* b ^* space, the color conversion table , The color signals of the points P1 to P4 are converted into the color signals of the points Q1 to Q4 in the CMY space, and then it is determined whether the points P1 to P4 belong to the color gamut or the color gamut. In the CMY space, since the range in the color gamut is distributed in a cubic range, it can be easily determined whether the color gamut is out of the color gamut.

  This conversion shows that the CMY values of the points P3 and P4 are registered in the color conversion table, but the CMY values of the points P1 and P2 are not registered in the color conversion table. If the CMY values of the points P1 and P2 are not registered in the color conversion table, it is difficult to obtain the CMY values corresponding to the point P by interpolation.

  Therefore, by reducing the saturation of the color signal at the point P, the point P is moved to the point P ′, and points P3 to P6 surrounding the point P ′ are extracted. Then, it is checked whether or not the CMY values of the points Q3 to Q6 corresponding to the points P3 to P6 are registered in the color conversion table, and the CMY values of the points Q3 to Q6 are registered in the color conversion table. Then, the CMY values of the point Q corresponding to the point P ′ are calculated by performing interpolation using the CMY values of the points Q3 to Q6.

  As a result of calculating the CMY value of the point Q, if the CMY value of the point Q belongs to the outside of the color gamut, a point S in the CMY space corresponding to the point R having the same luminance and achromatic color as the point P is obtained. Among the CMY numbers heading in the direction of the point S, the CMY number is replaced with the CMY value of the point Q ′ on the color gamut boundary, and then output to the printer.

FIG. 28 is a diagram for explaining exception processing of the method of FIG.
In Figure 28, the L ^* a ^* b ^* value to be converted, L ^* a ^* b ^* value of the point P significant color gamut is assumed entered. Here, assuming that the luminance of the color signal of the point P to be converted is extremely large or extremely small, the range of the color gamut in which the luminance is high or the luminance is low is narrow. No matter how much the saturation of the color signal is reduced, it is not possible to find the moving point surrounded by the grid points registered in the color conversion table. For this reason, for points where color conversion is impossible only by reducing the saturation, the color conversion is performed after the luminance is brought close to an intermediate value. When the brightness approaches an intermediate value, the range of the color gamut that can be reproduced is widened, so even if a point outside the large color gamut with extremely high brightness is input as the conversion target, conversion into the color gamut is performed. Can be done.

FIG. 29 is a block diagram showing the configuration of the color conversion apparatus according to the first embodiment of the present invention.
29, the color conversion device 61 includes an interpolation unit 62, a color gamut determination unit 63, and a replacement unit 64. The interpolation means 62 outputs the conversion result of the color signal by performing interpolation on the color signal to be converted. The color gamut determining means 63 determines whether the color signal conversion result belongs to the color gamut or the color gamut. When the conversion result of the color signal belongs to outside the color gamut, the replacement unit 64 replaces the color signal outside the color gamut obtained by the conversion with the color signal within the color gamut.

  As a result, it is possible to prevent the color signal outside the color gamut obtained by the conversion from being output as it is, and to prevent the handling of the color signal by the printer.

FIG. 30 is a block diagram showing the configuration of the color conversion apparatus according to the second embodiment of the present invention.
In FIG. 30, there are the following three types of L ^* a ^* b ^* grid points in the color conversion table 71, and CMY values are registered as follows.

1) In-gamut grid points (all CMY values are in the range of 0-255)
2) Out-of-gamut grid point (any CMY value is outside the range of 0-255)
3) Significantly out-of-gamut grid points (points without CMY values set)
The interpolation calculation unit 73 performs an interpolation calculation on the input L ^* a ^* b ^* value by referring to the color conversion table 71 using L ^* a ^* b ^* grid point data. Then, the color conversion processing unit 72 outputs the following three types of data for this interpolation calculation result.

4) Within the color gamut (all CMY values of the interpolation calculation result are within the range of 0 to 255)
5) Out of gamut (one of the CMY values of the interpolation operation result is outside the range of 0 to 255)
6) Significantly out of color gamut (interpolation calculation is not possible near the CMY value unset point (almost no normal image))
The color replacement processing unit 74 performs the following processing according to these three types of data.

7) In the color gamut: There is no additional processing performed by the color replacement processing unit 74, and the CMY values obtained by the interpolation calculation unit 73 are output as they are.
8) Out of color gamut: CMY values outside the color gamut are output to the color replacement unit 76, where they are changed to CMY values at the color gamut boundary (just below the color gamut) (color replacement processing). In this color replacement process, basically, a process of storing luminance and hue with CMY color signals is performed. For this reason, the interpolation calculation unit 75 refers to the color conversion table 71 to determine the CMY value corresponding to the achromatic value (L, 0, 0) having the same luminance as the L ^* a ^* b ^* value to be converted. The achromatic CMY value corresponding to this value (L, 0, 0) is output to the color replacement unit 76.

9) Out of large color gamut: The L ^* a ^* b ^* value input to the interpolation calculation unit 73 is input to the L ^* a ^* b ^* value changing unit 77. L ^* a ^* b ^* value change unit 77 changes the L ^* a ^* b ^* value while preserving the brightness and hue of the input to the interpolation operation unit 73 L ^* a ^* b ^* values, the interpolation operation unit 78 Output to. Interpolation operation unit 78, by performing a reference while interpolating the color conversion table 71, and calculates a CMY value corresponding to the L ^* a ^* b ^* is changed by a value changing section 77 L ^* a ^* b ^* values. In this interpolation calculation, when data indicating a color gamut or a significant color gamut is output, the L ^* a ^* b ^* value changing unit 77 further changes the L ^* a ^* b ^* value to the interpolation calculating unit 78. Output. When the interpolation calculation result is out of the color gamut, this CMY value is output to the color replacement unit 76.

  Note that this processing outside of the color gamut requires repeated processing until a value outside the color gamut is found, so it takes some processing time, but it is not necessary to calculate the value accurately (the color outside the color gamut has a certain width). Therefore, it is possible to perform by rough processing. For this reason, it is possible to perform the arithmetic processing at a higher speed than in the case where the CMY value of the color near the color gamut is obtained by repeated arithmetic.

Hereinafter, the interpolation calculation process will be described more specifically using mathematical expressions.
Let the input L ^* a ^* b ^* value be (Li, ai, bi). Further, since the CMY values corresponding to the L ^* a ^* b ^* numbers (L, a, b) of the grid points are registered in the color conversion table, the actual creation data format is C [L] [a] [ b], M [L] [a] [b], Y [L] [a] [b]. Hereinafter, description will be made assuming that the interpolation calculation processing is performed by cubic interpolation.

First, lattice points used for interpolation (eight points surrounding L ^* a ^* b ^* values to be converted) are selected by the following expression.
L ^* = (int) (Li / g) (125)
a ^* = (int) ((ai + 1128.0) / g) (126)
b ^* = (int) ((bi + 128.0) / g) (127)
Here, g represents the lattice size, and (int) represents the decimal point truncation.

Next, assuming that the length w of the side of the cube is 1, the position (Lw, aw, bw) of the L ^* a ^* b ^* value inside the cube is calculated by the following formula.
Lw = (Li / g−L) (128)
aw = ((ai + 1128.0) / ga) (129)
bw = ((bi + 128.0) / g−b) (130)
Next, the weighting factor (volume (V (0,0,0) to V (1,1,1)) generated by dividing) used for cube interpolation is calculated. Here, the volume (V (0, 0, 0) to V (1, 1, 1) can be calculated from the position (Lw, aw, bw) of the L ^* a ^* b ^* value inside the cube.

  FIG. 31 is a diagram illustrating an example of eight rectangular parallelepipeds divided at positions (Lw, aw, bw) inside a cube. From FIG. 31, the volume (V (0,0,0) to V (1,1,1)) of the rectangular parallelepiped can be easily calculated as shown by the following equation.

V (0,0,0) = (1-Lw). (1-aw). (1-bw)
... (131)
V (1, 0, 0) = Lw. (1-aw). (1-bw) (132)
V (0,1,0) = (1-Lw) · aw · (1-bw) (133)
V (0,0,1) = (1-Lw). (1-aw) .bw (134)
V (0,1,1) = (1-Lw) · aw · bw (135)
V (1, 0, 1) = Lw · (1-aw) · bw (136)
V (1,1,0) = Lw · aw · (1-bw) (137)
V (1,1,1) = Lw · aw · bw (138)
Next, it is confirmed whether there are any unset points among the 8 points surrounding the position (Lw, aw, bw) inside the cube. Not performed. Here, when the following conditions are satisfied, there is an unset point.

(A) L, a, b, L + 1, a + 1, b + 1 are out of the existence range of the color conversion table (out of the secured range of the three-dimensional arrangement of the color conversion table)
(B) Any negative of Lw, aw, bw (c) L ^* a ^* b ^* number is (L, a, b), (L + 1, a, b), (L, a + 1, b), (L, There are lattice points in which no data is set in (a, b + 1), (L, a + 1, b + 1), (L + 1, a, b + 1), (L + 1, a + 1, b), (L + 1, a + 1, b + 1).

Next, the L ^* a ^* b ^* numbers are (L, a, b), (L + 1, a, b), (L, a + 1, b), (L, a, b + 1), (L, a + 1, b + 1). , (L + 1, a, b + 1), (L + 1, a + 1, b), (L + 1, a + 1, b + 1) to the CMY values of V (0, 0, 0) to V (1, 1, 1) Interpolation processing is executed by calculating the weighted average.

  Here, C [L] [a] [b], M [L] [a] [b], and Y [L] [a] [b] are the CMY values of the lattice point (L, a, b), respectively. , CMY values to be interpolated can be calculated by the following equation.

C = (C [L] [a] [b] · V (0,0,0)
+ C [L + 1] [a] [b] · V (1, 0, 0)
+ C [L] [a + 1] [b] · V (0,1,0)
+ C [L] [a] [b + 1] · V (0, 0, 1)
+ C [L] [a + 1] [b + 1] · V (0,1,1)
+ C [L + 1] [a] [b + 1] · V (1, 0, 1)
+ C [L + 1] [a + 1] [b] · V (1,1,0)
+ C [L + 1] [a + 1] [b + 1] · V (1,1,1))
... (139)
M = (M [L] [a] [b] · V (0,0,0)
+ M [L + 1] [a] [b] · V (1, 0, 0)
+ M [L] [a + 1] [b] · V (0,1,0)
+ M [L] [a] [b + 1] · V (0, 0, 1)
+ M [L] [a + 1] [b + 1] · V (0,1,1)
+ M [L + 1] [a] [b + 1] · V (1, 0, 1)
+ M [L + 1] [a + 1] [b] · V (1,1,0)
+ M [L + 1] [a + 1] [b + 1] · V (1,1,1))
... (140)
Y = (Y [L] [a] [b] · V (0, 0, 0)
+ Y [L + 1] [a] [b] · V (1, 0, 0)
+ Y [L] [a + 1] [b] · V (0,1,0)
+ Y [L] [a] [b + 1] · V (0, 0, 1)
+ Y [L] [a + 1] [b + 1] · V (0,1,1)
+ Y [L + 1] [a] [b + 1] · V (1, 0, 1)
+ Y [L + 1] [a + 1] [b] · V (1,1,0)
+ Y [L + 1] [a + 1] [b + 1] · V (1,1,1))
... (141)
Next, the CMY values calculated by the equations (139) to (141) are output. If any of the CMY values does not fall within the range of 0 to 255, it is out of the color gamut.

  Next, the color replacement process will be described more specifically using mathematical expressions. First, before executing the color replacement process, calculation of the maximum luminance value Lmax and the minimum luminance value Lmin of the achromatic color is performed in advance.

To calculate the luminance maximum value Lmax of achromatic color, while increasing the L ^* a ^* b ^* value (50, 0, 0) and L ^* by one, performs color conversion to the CMY value, the color conversion result The L ^* value that is out of the color gamut or significantly out of the color gamut is set to a value smaller by one.

In order to calculate the maximum luminance value Lmax of the achromatic color, the L ^* a ^* b ^* value is reduced from (50, 0, 0) by 1 while the L ^* value is reduced by 1, and the color conversion result is obtained. Is a value that is one greater than the L ^* value that is out of the gamut or significantly out of the gamut.

Next, it is determined whether the interpolation calculation result is “out of gamut” or “out of gamut”.
Hereinafter, a case where the interpolation calculation result is determined to be “out of color gamut” will be described.
First, the CMY value (Cn, Mn, Yn) of the same luminance achromatic color (Li, 0, 0) is calculated by using the interpolation calculation process.

In this interpolation calculation calculation, the Li value is clipped in the range of (Lmax−Lmargin) to (Lmin + Lmargin). That is,
When Li> Lmax-Lmargin,
Li = Lmax−Lmargin,
When Li <Lmin + Lmargin,
Let Li = Lmin + Lmargin.

However, Lmargin is a predetermined constant. Here, the Li value is clipped in the range of (Lmax−Lmargin) to (Lmin + Lmargin) because the color gamut is extremely narrow in the vicinity of the minimum luminance value Lmin and the maximum luminance value Lmax when viewed from the a ^* b ^* plane. This is because there are many cases.

In the subsequent processing, basically, the CMY value just near the color gamut between the CMY value of the same luminance achromatic color and the CMY value outside the color gamut is obtained, but the color gamut on the a ^* b ^* plane is narrow (no If there is only a color close to a chromatic color), the obtained CMY values are biased to colors close to an achromatic color. Therefore, a value that is extremely close to the vicinity of the minimum luminance value Lmin and the maximum luminance value Lmax is not used as the Li value of the same luminance achromatic color.

Next, among the CMY components exceeding the color reproduction range, the component exceeding the “far” is determined, and the excess level D is calculated.
FIG. 32 is a diagram for explaining the distance level of the CMY component beyond the color reproduction range.

  In FIG. 32, when the CMY value is less than 0, the difference d1 between the achromatic CMY value and the CMY value 0 and the difference d2 between the achromatic CMY value and the out-of-gamut CMY value are obtained. At this time, the distance level is evaluated by d2 / d1.

  On the other hand, if the CMY value exceeds 255, the difference d3 between the achromatic CMY value and the CMY value 255 and the difference d4 between the achromatic CMY value and the out-of-gamut CMY value are obtained. At this time, the distance level is evaluated by d4 / d3.

Specifically, the distance level (Cd, Md, Yd) is calculated by the following equation.
Cd = 0, Md = 0, Yd = 0 (initialization)
if (C <0) Cd = (C-Cn) / (0-Cn) (142)
if (C> 255) Cd = (Cn-C) / (Cn-255)
... (143)
if (M <0) Md = (M−Mn) / (0−Mn) (144)
if (M> 255) Md = (Mn-M) / (Mn-255)
... (145)
if (Y <0) Yd = (Y−Yn) / (0−Yn) (146)
if (Y> 255) Yd = (Yn−Y) / (Yn−255)
... (147)
The maximum level among these distant levels (Md, Cd, Yd) is defined as an excess level D.

Next, the CMY value (Cb, Mb, Yb) of the color just in the color reproduction range is calculated by the following formula.
Cb = Cn− (Cn−C) / D (148)
Mb = Mn- (Mn-M) / D (149)
Yb = Yn− (Yn−Y) / D (150)
Next, a case where the interpolation calculation result is determined to be “outside the large color gamut” will be described. In this case, the input L ^* a ^* b ^* value is shifted from “outside color gamut” to “outside color gamut” while satisfying the equi-luminance equal hue. That is, the process of reducing the saturation of the input L ^* a ^* b ^* value is performed.

Specifically, first, the initial values of the saturation crit and gdiv of the input L ^* a ^* b ^* values (Li, ai, bi) are set by the following equations.
crinit = cr = sqrt (ai · ai + bi · bi) (saturation setting)
... (151)
gdiv = gdivinit (preset initial value) (152)
Next, when the saturation crit of the input L ^* a ^* b ^* value (Li, ai, bi) is reduced by the following equation and the saturation cr obtained as a result of the reduction is smaller than 0, Perform exception handling.

cr = cr-g / gdiv (153)
However, g is a lattice interval of L ^* a ^* b ^* space.
Next, an a ^* b ^* value (anew, bnew) obtained as a result of reducing the saturation is obtained by the following equation.

anew = ai · (cr / crinit) (154)
bnew = bi · (cr / crinit) (155)
Next, it is determined whether the L ^* a ^* b ^* value (Li, anew, bnew) obtained as a result of reducing the saturation is out of the color gamut, and this L ^* a ^* b ^* value (Li, anew, bnew). ) Is out of the color gamut, color conversion processing is executed for the L ^* a ^* b ^* values (Li, anew, bnew). On the other hand, when the L ^* a ^* b ^* value (Li, anew, bnew) is outside the large color gamut, a process for further reducing the saturation is performed according to the equation (153). Also, when the L ^* a ^* b ^* values (Li, anew, bnew) are within the color gamut, a process for increasing the saturation is performed according to the following equation.

gdiv = gdiv + gdiv (the recommended value of the maximum value of gdiv is 32)
... (156)
cr = cr + g / gdiv (157)
Then, the a ^* b ^* value (anew, bnew) obtained as a result of increasing the saturation is obtained by the equations (154) and (155), and the L ^* a ^* b ^* value (Li, anew, bnew) is obtained. Continue to change saturation until it goes out of gamut.

Next, exception processing will be described.
In this exceptional process, the L ^* a ^* b ^* value of the same luminance achromatic color of the input L ^* a ^* b ^* value is moved from “out of the large color gamut” to “out of the color gamut”. That is, the process of bringing the luminance of the input L ^* a ^* b ^* value closer to 50 is performed.

Specifically, first, the L ^* a ^* b ^* value (Li, anew, bnew) of the same luminance achromatic color of the input L ^* a ^* b ^* value (Li, ai, bi), the initial value of gdiv, and The initial value of Lref is set by the following formula.

gdiv = dgivint (recommended value is 2) (158)
Lrefinit = Lref = abs (Li-50) (159)
anew = bnew = 0 (160)
Here, abs represents an absolute value calculation.

Next, a new L ^* a ^* b ^* value (Lnew, 0, 0) is generated by bringing the luminance closer to 50 using the following equation.
Lref = Lref−g / gdiv (161)
Lnew = 50 + (Li-50). (Lref / Lrefinit)
... (162)
However, g is a lattice interval of L ^* a ^* b ^* space.

Next, it is determined whether or not the newly obtained L ^* a ^* b ^* value (Lnew, 0, 0) is out of the color gamut, and this L ^* a ^* b ^* value (Lnew, 0, 0) is out of the color gamut. In this case, a color conversion process is executed for this L ^* a ^* b ^* value (Lnew, 0, 0). On the other hand, when the L ^* a ^* b ^* value (Lnew, 0, 0) is significantly outside the color gamut, processing for further bringing the luminance closer to 50 is performed according to equation (162). Further, when the L ^* a ^* b ^* value (Lnew, 0, 0) is in the color gamut, processing for moving the luminance away from 50 is performed by the following formula.

gdiv = gdiv + gdiv (the recommended value of the maximum value of gdiv is 32)
... (163)
Lref = Lref + g / gdiv (164)
Lnew = 50 + (Li-50). (Lref / Lrefinit)
... (165)
Then, the L ^* value Lnew obtained as a result of moving the luminance away from 50 is obtained by the equation (162), and the luminance is continuously changed until the L ^* a ^* b ^* value (Lnew, 0, 0) falls outside the color gamut. .

FIG. 33 is a flowchart showing color conversion processing according to an embodiment of the present invention.
In FIG. 33, the color signal to be converted is converted to a color signal in a different color space by performing interpolation using the color signal registered in the color conversion table (step S91).

  Next, it is determined whether or not the conversion result belongs to the color gamut of the color reproduction device (step S92). If the conversion result does not belong to the color gamut of the color reproduction device, the conversion result is within the color gamut of the color reproduction device. The color signal is replaced and output (step S93).

FIG. 34 is a flowchart showing an interpolation process according to an embodiment of the present invention.
In FIG. 34, first, when an L ^* a ^* b ^* value to be converted is input (step S101), L of lattice points used for interpolation (8 points surrounding the L ^* a ^* b ^* value to be converted). ^* a ^* b ^* number is set (step S102).

Next, for the L ^* a ^* b ^* value to be converted, the position inside the cube whose vertex is the lattice point used for interpolation is calculated (step S103), and the cube is divided at the calculated point By calculating the volume of the rectangular parallelepiped, a weighting coefficient used for interpolation is calculated (step S104).

Next, whether or not CMY values corresponding to L ^* a ^* b ^* grid points used for interpolation are registered is confirmed by referring to the color conversion table (step S105), and eight L ^* used for interpolation are checked ^. Some a ^* b ^* grid points is determined, if the L ^* a ^* b ^* grid points CMY value is not registered exists even one with significant color gamut (step S106).

On the other hand, when CMY values are registered in the color conversion table for all eight L ^* a ^* b ^* grid points used for interpolation, the CMY values are multiplied by the weighting coefficients calculated in step S104. Then, an interpolation calculation is performed (step S107).

  Next, it is determined to which range in the CMY space the CMY values obtained as a result of the interpolation calculation belong (step S108). If all the CMY values obtained as a result of the interpolation calculation are within the range of 0 to 255, it is determined that the color gamut is within the range, and the CMY value is output (step S109).

  On the other hand, if any of the CMY values obtained as a result of the interpolation calculation is outside the range of 0 to 255, it is determined that the color gamut is outside, and the CMY value is replaced with a CMY value within the color gamut (step S110). .

FIG. 35 is a flowchart showing color replacement processing according to an embodiment of the present invention.
In FIG. 35, first, when an interpolation calculation result is input (step S121), it is determined whether the interpolation calculation result is out of the color gamut or out of the large color gamut (step S122).

  Then, when it is determined that the color gamut is outside, CMY values having the same luminance as the CMY values outside the color gamut (Cn, Mn, Yn) are calculated (step S123), and the equations (142) to (147) are used. An excess level indicating how far the CMY values outside the color gamut are away from the color gamut boundary is calculated (step S124).

  Next, by moving the out-of-gamut CMY values in the direction of the same luminance achromatic CMY values (Cn, Mn, Yn) by the excess level, the CMY values (Cb, Mb, Yb) is calculated (step S125). Then, this CMY value (Cb, Mb, Yb) is output as a mapping result (step S126).

On the other hand, if it is determined in step S122 that it is out of the color gamut, the saturation of the input L ^* a ^* b ^* value and the change range of the saturation are set (step S127), and the input L ^* a ^* The saturation of the b ^* value is decreased (step S128). If the saturation is less than 0 as a result of reducing the saturation (step S129), exception processing is performed (step S130).

On the other hand, if the saturation is 0 or more as a result of decreasing the saturation, the a ^* b ^* value is changed (step S131). Then, the a ^* b ^* result value changing the resulting L ^* a ^* b ^* values, whether CMY values of the vertices of the cube containing the L ^* a ^* b ^* values are registered in all the color conversion table To determine whether the changed L ^* a ^* b ^* value is out of the color gamut (step S132). As a result, if the changed L ^* a ^* b ^* value is outside the color gamut, the process proceeds to step S123, and the CMY value outside the color gamut is replaced with the CMY value within the color gamut.

On the other hand, if the L ^* a ^* b ^* value after the change is not out of the color gamut, it is determined whether or not the L ^* a ^* b ^* value after the change is out of the large color gamut (step S133) ^. If the a ^* b ^* value is significantly outside the color gamut, the process returns to step S128 to further reduce the saturation of the L ^* a ^* b ^* value.

On the other hand, if the changed L ^* a ^* b ^* value is within the color gamut, the saturation change range set in step S127 is reduced (step S134), and the saturation of the L ^* a ^* b ^* value Is increased (step S135), and the a ^* b ^* value is changed (step S136). Then, the a ^* b ^* result value changing the resulting L ^* a ^* b ^* values, whether CMY values of the vertices of the cube containing the L ^* a ^* b ^* values are registered in all the color conversion table Is checked to determine whether the changed L ^* a ^* b ^* value is out of the color gamut (step S137). As a result, if the changed L ^* a ^* b ^* value is outside the color gamut, the process proceeds to step S123, and the CMY value outside the color gamut is replaced with the CMY value within the color gamut.

On the other hand, if the L ^* a ^* b ^* value after the change is not out of the color gamut, it is determined whether or not the L ^* a ^* b ^* value after the change is out of the large color gamut (step S138), and the L ^* after the change is determined ^. If a ^* b ^* value is significantly color gamut, the flow returns to step S128, the L ^* a ^* b ^* values saturation reduces the, if the L ^* a ^* b ^* value after the change in color gamut, Returning to step S134, the saturation change width of the L ^* a ^* b ^* value is further reduced.

FIG. 36 is a flowchart showing exception processing according to an embodiment of the present invention.
In Figure 36, the input L ^* a ^* b ^* values of a ^* b ^* values by zero, the inputted L ^* a ^* b ^* of the luminance achromatic values L ^* a ^* b ^* value Is set (step S141).

Next, an L ^* a ^* b ^* value luminance change range is set (step S142), and the luminance is brought close to 50 (step S143), thereby changing the L ^* a ^* b ^* value.
Next, it is determined whether or not the L ^* a ^* b ^* value after the change is out of the color gamut (step S144). If the L ^* a ^* b ^* value after the change is out of the color gamut, the process proceeds to step S139 in the flowchart of FIG. Proceed (step S150).

On the other hand, if the L ^* a ^* b ^* value after the change is not out of the color gamut, it is determined whether or not the L ^* a ^* b ^* value after the change is out of the large color gamut (step S145), and the L ^* a after the change is determined. ^{If the *} b ^* value is significantly outside the color gamut, the process returns to step S143, and the luminance of the changed L ^* a ^* b ^* value is further brought closer to 50.

On the other hand, if the changed L ^* a ^* b ^* value is in the color gamut, the luminance change width set in step S142 is reduced (step S146), and the changed L ^* a ^* b ^* value The brightness is moved away from 50 (step S147). Then, for L ^* a ^* b ^* values obtained as a result of moving the brightness away from 50, it is checked whether all CMY values of the vertices of the cube including the L ^* a ^* b ^* values are registered in the color conversion table. Thus, it is determined whether or not the changed L ^* a ^* b ^* value is out of the color gamut (step S148). As a result, if the L ^* a ^* b ^* value after the change is out of the color gamut, the process proceeds to step S150, and the processes after step S139 in the flowchart of FIG. 35 are performed.

On the other hand, it is determined whether the L ^* a ^* b ^* value after the change not in the gamut, the L ^* a ^* b ^* value after the change is significantly color gamut (step S149), the changed L ^* If a ^* b ^* value is significantly color gamut, the flow returns to step S143, L ^* a ^* b ^* luminance further closer to 50 the value, if the L ^* a ^* b ^* value after the change in color gamut Returning to step S146, the luminance change width of the L ^* a ^* b ^* value is further reduced.

  In this way, by replacing a color signal outside the color gamut with a color signal within the color gamut, it is possible to prevent a value outside the color reproduction range of the device from being output when performing color conversion, The color signal to be calculated can be calculated at high speed.

In the embodiment of FIG. 30, the color conversion apparatus has been described. However, the interpolation calculation unit 73 performs an interpolation calculation and outputs the calculation result (CMY value), whereby the input L ^* a ^* b ^* value. Can be applied to a color gamut determination device that determines whether the color gamut is within or outside the color gamut of the device.

In particular, if an interpolation operation is performed for the purpose of determining whether the L ^* a ^* b ^* value is within or outside the device gamut, an identifier indicating the color gamut boundary, an identifier indicating the color gamut inside, and an identifier indicating the color gamut outside May be added to the color conversion table.

Specifically, a certain L ^* a ^* b ^* lattice point (L, a, b) and surrounding (L + 1, a, b), (L, a + 1, b), (L, a, b + 1), All L ^* a ^* b ^* grid points of (L, a + 1, b + 1), (L + 1, a, b + 1), (L + 1, a + 1, b), (L + 1, a + 1, b + 1) are CMY values in the device gamut. Is added to the L ^* a ^* b ^* grid point, an identifier indicating the color gamut.

Further, when a certain L ^* a ^* b ^* grid point and all surrounding L ^* a ^* b ^* grid points correspond to CMY values (or not set) outside the device color gamut, the L ^* a ^* b ^* An identifier indicating the out-of-gamut is added to the grid point.

Also, when a certain L ^* a ^* b ^* grid point and a part of the surrounding L ^* a ^* b ^* grid points correspond to CMY values outside the device gamut (or not set), the gamut boundary Set the identifier.

By registering such an identifier in the color conversion table, when an L ^* a ^* b ^* value to be converted is input, prior to the interpolation operation for the L ^* a ^* b ^* value, the conversion target For most of the L ^* a ^* b ^* values, it can be determined at high speed whether it is in the color gamut or outside the color gamut. Then, by actually performing the interpolation operation only on the L ^* a ^* b ^* value whose identifier indicates the color gamut boundary, whether all the conversion target L ^* a ^* b ^* values are within the color gamut or out of the color gamut. Can be determined.

In particular, if the color conversion table is used only for the determination of inside and outside of the color gamut, it is not necessary to set CMY values for L ^* a ^* b ^* grid points to which identifiers within and outside the color gamut are added. Is possible.

Further, as an identifier, an identifier indicating the inside or outside of the color gamut may be added based on only the CMY value of a certain grid point (L, a, b).
FIG. 37 is a diagram showing an example of the content of the color conversion table when an identifier indicating whether or not the L ^* a ^* b ^* value is in the color gamut is added.

In FIG. 37, assuming that the CMY values that can be reproduced by the printer are in the range of 0 to 255, the L ^* a ^* b ^* values are (0, 0, 0), (48, 0, -96), ( 48, 0, −80) L ^* a ^* b ^* grid points are assigned identifiers indicating out of the color gamut, and L ^* a ^* b ^* values are (16, 0, 0), (32, 0). , 0), (48, 0, −64), (48, 0, −48), and (48, 0, −32) L ^* a ^* b ^* grid points have identifiers indicating the color gamut. It has been added.

Then, when the L ^* a ^* b ^* value to be converted is input, the following L ^* a ^* b ^* grid point identifiers are referred to prior to the interpolation calculation for the L ^* a ^* b ^* value.
(L, a, b), (L + 1, a, b), (L, a + 1, b),
(L, a, b + 1), (L, a + 1, b + 1), (L + 1, a, b + 1),
(L + 1, a + 1, b), (L + 1, a + 1, b + 1)
When all the identifiers added to these L ^* a ^* b ^* grid points indicate the color gamut, it is determined that the L ^* a ^* b ^* values to be converted belong to the color gamut, and these L ^* a When all identifiers added to ^* b ^* grid points indicate out of gamut, it is determined that the conversion target L ^* a ^* b ^* values belong to outside the gamut, and are added to these L ^* a ^* b ^* grid points. In the case where there are identifiers indicating the color gamut and identifiers indicating the color gamut, the L ^* a ^* b ^* value to be converted belongs to the color gamut or falls outside the color gamut by actually performing an interpolation operation. Determine if it belongs.

In such a color gamut determination, when a color conversion table without an identifier is used, the following CMY values of L ^* a ^* b ^* lattice points are examined in advance.
(L, a, b), (L + 1, a, b), (L, a + 1, b),
(L, a, b + 1), (L, a + 1, b + 1), (L + 1, a, b + 1),
(L + 1, a + 1, b), (L + 1, a + 1, b + 1)
When all the CMY values of these L ^* a ^* b ^* grid points are in the range of 0 to 255, it is determined that the L ^* a ^* b ^* values to be converted belong to the color gamut, and these When all the CMY values of the L ^* a ^* b ^* grid points are less than 0 or more than 255, it is determined that the L ^* a ^* b ^* values to be converted belong outside the color gamut, and these L ^* a ^* b ^* If some of the CMY values at the grid points are less than 0 or greater than 255, whether the L ^* a ^* b ^* value to be converted belongs to the color gamut or the color gamut by actually performing an interpolation operation. Determine.

As mentioned above, although the Example of this invention was described, this invention is not limited to the Example mentioned above, Various other changes are possible within the range of the technical idea of this invention.
For example, in the above-described embodiment, the case where the L ^* a ^* b ^* signal is converted into the CMY signal has been described. However, in addition to this case, the present invention can be applied to an RBG signal or an XYZ signal. It can also be adapted to CMYK (cyan, magenta, yellow, black) signals.

It is a block diagram which shows the structure of the color conversion table concerning one Example of this invention. It is a figure which shows the correspondence of the L ^* a ^* b ^* signal and CMY signal concerning one Example of this invention on a two-dimensional color space. It is a table | surface which shows the numerical example registered into the color conversion table concerning one Example of this invention. 1 is a block diagram illustrating a configuration of a color signal generation apparatus according to an embodiment of the present invention. 3 is a flowchart illustrating a color signal generation method according to an embodiment of the present invention. 1 is a block diagram showing a system configuration of a color signal generation apparatus according to an embodiment of the present invention. It is the figure which showed on the color space the color signal generation method concerning 1st Example of this invention. (A) is a figure which shows cubic interpolation, (b) is a figure which shows triangular prism interpolation, (c) is a figure which shows tetrahedral interpolation. It is a figure which shows the correspondence at the division | segmentation point of the CMY signal and L ^* a ^* b ^* signal concerning one Example of this invention on a two-dimensional color space. It is a figure which shows the structure of the color conversion table production | generation apparatus concerning one Example of this invention. It is a flowchart which shows the data selection process concerning one Example of this invention. It is a flowchart which shows the interpolation calculation process and data registration process concerning one Example of this invention. It is a figure which shows the extrapolation point selection process concerning one Example of this invention two-dimensionally on L ^* a ^* b ^* space. It is a figure which shows the extrapolation point production | generation method concerning one Example of this invention two-dimensionally on color space. It is a figure which shows the adjacent point on the three-dimensional space concerning one Example of this invention. 3 is a flowchart illustrating a method for generating a color signal outside a color gamut according to the first embodiment of the present invention. It is a table | surface which shows the content of the color conversion table produced | generated by the process of FIG. It is the figure which showed on the color space the color signal generation method concerning 2nd Example of this invention. It is a figure which shows the extrapolation direction in the color signal production | generation method concerning 2nd Example of this invention. It is a figure which shows the calculation method of the adjacent point used with the color signal generation method concerning 2nd Example of this invention. It is a figure which shows the extrapolation distance calculation method in the color signal generation method concerning 2nd Example of this invention. It is a flowchart which shows the color signal generation method concerning 2nd Example of this invention. It is a flowchart which shows the color signal production | generation method outside a color gamut concerning 2nd Example of this invention. It is a flowchart which shows the color signal generation method corresponding to the L ^* a ^* b ^* lattice point concerning 2nd Example of this invention. It is a figure which shows the color conversion table unset point concerning one Example of this invention on a color space. It is a figure which shows the interpolation method of the color signal outside a color gamut concerning one Example of this invention on a two-dimensional color space. It is a figure which shows the interpolation method of the color signal outside a large color gamut concerning one Example of this invention on a two-dimensional color space. It is a figure explaining the exception process of the color signal outside a large color gamut concerning one Example of this invention on a two-dimensional color space. 1 is a block diagram illustrating a configuration of a color conversion apparatus according to a first embodiment of the present invention. It is a block diagram which shows the structure of the color conversion apparatus concerning 2nd Example of this invention. It is a figure explaining the calculation method of the weighting coefficient used by the cube interpolation concerning one Example of this invention. It is a figure explaining the excess level of the CMY signal outside a color gamut concerning one Example of this invention. 3 is a flowchart illustrating a color conversion method according to an embodiment of the present invention. 3 is a flowchart illustrating a color gamut determination method according to an embodiment of the present invention. It is a flowchart which shows the color replacement process concerning one Example of this invention. It is a flowchart which shows the exception process concerning one Example of this invention. It is a figure which shows the content of the color conversion table which registered the identifier for color gamut determination concerning one Example of this invention. It is a figure which shows the correspondence of the conventional L ^* a ^* b ^* signal and CMY signal on a two-dimensional color space. It is a figure which shows the content of the conventional color conversion table.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1, 43, 52, 71 Color conversion table 2 1st color conversion result 3 2nd color conversion result 11 Color signal generation apparatus 12 Color signal selection means 13 Color signal extrapolation means 21 CPU
22 ROM
23 RAM
24 Input / Output Interface 25 Display 26, 41 Printer 27 Memory 28 Scanner 29 Communication Interface 30 Communication Network 31 Driver 32 Hard Disk 33 IC Memory Card 34 Magnetic Tape 35 Flexible Disk 36 Optical Disk 37 Bus 42 Colorimeter 44, 54 CMY Numbers 45, 53 L ^* a ^* b ^* No. 46 CMY value calculation unit 47 Interpolation unit 48 Proximity detection unit 49 Proximity calculation unit 50 Comparison selection unit 51 Registration unit 55 Proximity 61 Color conversion device 62 Interpolation unit 63 Color gamut determination unit 64 Replacement unit 72 Color conversion processing unit 73, 75, 78 Interpolation calculation unit 74 Color replacement processing unit 76 Color replacement unit 77 L ^* a ^* b ^* Value change unit

Claims (1)

Input means for inputting a color signal of a predetermined first color space;
A table for registering a color signal conversion result to a second color space expressed in a representation format depending on a predetermined device with respect to an input of a color signal of the first color space, the conversion result being For a range exceeding the color reproducible range in the predetermined device, a color conversion table described by color signal values that cannot be reproduced by the device, and
Color signal selection means for selecting a color signal of a first color space registered in the color conversion table surrounding the input color signal;
Detecting means for detecting a color signal in the second color space corresponding to the selected color signal by referring to the color conversion table;
A weight calculating means for calculating a weighting coefficient based on the position of the input color signal in the first color space;
Based on the color signal in the second color space detected from the color conversion table and the weighting factor, the second color space corresponding to the color signal in the first color space input from the input means. First color signal calculation means for calculating a color signal ;
First color gamut determining means for determining whether the color signal of the second color space calculated by the first color signal calculating means is within the color reproduction range of the device;
When the first color gamut determination unit determines that the color reproduction range of the device is exceeded, the color signal calculated by the first color signal calculation unit is changed to a color within the color reproduction range of the device. First color replacement means for replacing with a signal;
A color conversion device comprising:

Images