JP2000341545A - Device and method for proof-reading data table - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain data table proof-reading device and method capable of more easily calibrating a data table. SOLUTION: The flow of data table calibration includes first, calculating 2(n-2)*2(n-2)*2(n-2) pieces of difference data from 2n*2n*2n pieces of output data and 2(n-2)*2(n-2)*2(n-2) pieces of newly calculated update data (S401). Next, 2n*2n*2n pieces of difference data are calculated for the whole input data of an old data table by performing linear interpolation of the difference data (S403). Then, 2n*2n*2n pieces of new table data are calculated from these whole difference data and the old table data (S405). Thus, it is possible to calibrate the old data table by calculating the new table data which number as many as the old table data from update data being less than the old table data in the number of pieces of data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data table calibrating apparatus and method, and more particularly, to a data table calibrating apparatus and method for calibrating table data in an image processing apparatus using a look-up table for color conversion. About.

[0002]

2. Description of the Related Art In image output apparatuses such as digital copiers and printers, it is necessary to perform predetermined image processing on input data in order to obtain output data suitable for the device. For example, in a color copying machine, it is necessary to perform a color conversion process on RGB data, which is input data read by a reading unit, into CMY data or the like that can be represented by a color material of the color copying machine.

[0003] However, when this color conversion processing is performed by calculation, it is very complicated, and a considerable processing time is required.
Therefore, in general, some measures have been taken to shorten the processing time. For example, a method of using a look-up table in which output data for input data is registered in advance is employed.

FIG. 6 is a flowchart showing the flow of a general color conversion process using a look-up table. Referring to FIG. 6, first, in step S601, color information is input to the color conversion processing unit and temporarily stored. Here, the color information is three-dimensional color image data of one pixel, for example, RGB corresponding to red, green, and blue.
Each has an 8-bit value.

[0005] Next, in step S603, output color information (CM) corresponding to the RGB values that are the input color information.
A reference address of a lookup table indicating a storage location of (Y value) is set. Subsequently, in step S605, the output color information (CMY values) stored in the lookup table is referred to according to the set reference address.

The output color information (CMY values) of the lookup table referred to is temporarily stored as an output value in step S607, and output from the color conversion processing unit in step S609.

This series of processing is repeated by the number of color information, that is, by the number of pixels of the input image.

FIG. 7 shows a typical configuration of the lookup table referred to. As shown in FIG. 7, the look-up table stores the 8-bit values of RGB as input color information (described in the left half) and the 8-bit values of CMY corresponding to output color information (right half). ) Are stored. In addition, due to problems such as memory capacity,
Normally, the color space is divided into an appropriate number of grid points, and only the data of the grid points is tabulated.

Therefore, when the color information input to the color conversion processing unit matches the grid point, the CMY values corresponding to the RGB values, which are the color information, are referred to and output as output data. If the obtained color information exists between the lattice points, a linear interpolation operation is performed during the actual color conversion processing to obtain an output value. It should be noted that even if the linear interpolation operation is performed, the processing time is much shorter than the color conversion processing in which the calculation is performed without using the lookup table.

As described above, when the look-up table is used, the color conversion process itself requires only the output of the output color information corresponding to the input color information with reference to the memory or the output of the output color information by the linear interpolation operation. Simple processing is required. Therefore, once the lookup table is created, the color conversion processing can be performed very simply and at high speed.

However, creating this look-up table requires a large amount of colorimetric data, which requires considerable time and effort. For example, if an attempt is made to tabulate the data of all 8 bits of RGB, 24 bits, that is, 8 bits of CMY data for 2 ²⁴ points must be stored. By thinning out the lattice points, even if the performing interpolation calculation, also the data from the 2 ⁷ 2 ⁹ or so is required.

Moreover, the correspondence between the color information is not simple, and differs depending on the characteristics of the equipment used. Therefore, these data must be actually obtained by experiments.
Specifically, each time using a color chart or the like,
It is necessary to perform input / output and colorimetry with a device to be actually used, and obtain a correspondence using the colorimetric data.

As described above, it takes a lot of time and effort to acquire data for creating a look-up table, so that the look-up table once created cannot be easily calibrated. Therefore, even if the color reproduction is shifted from that at the time of creating the look-up table due to environmental changes, device fluctuations, and the like, it is difficult to create a new look-up table again as a practical problem.

Therefore, when it is necessary to change the look-up table, table data calibration such as re-creating the look-up table itself is not performed, and another color conversion processing method is adopted. Was normal.

[0015]

However, if another color conversion processing method is adopted without calibrating the table data, a problem arises that the color conversion processing takes time. Therefore, if table data can be easily calibrated, it is desired to employ a look-up table type color conversion process in which output data can be obtained immediately for input data.

The present invention has been conceived in view of such circumstances, and an object of the present invention is to provide a data table calibration method and apparatus which can more easily perform data table calibration.

[0017]

To achieve the above object, according to one aspect of the present invention, N pieces of input color information,
A data table calibrating apparatus for calibrating a data table for recording N output color information corresponding to the N input color information includes M (M <N) new input color information, and M (M <N) new input color information. An acquisition unit that acquires M pieces of new output color information corresponding to each of the new pieces of input color information, and M data corresponding to each of the M pieces of new input color information from the data table.
A reading unit that reads out the pieces of output color information, and calculates a difference between each of the read out pieces of M output color information and each of the obtained M pieces of new output color information as M pieces of difference data. A first calculating unit for calculating N difference data by interpolating the M difference data, and a second calculating unit for calculating the N difference data. A calibration unit for calibrating the N pieces of output color information.

According to the present invention, it is possible to provide a data table calibrating device capable of easily calibrating a data table.

Preferably, the second calculating means performs linear interpolation of the M pieces of difference data. This makes it possible to easily calculate all difference data.

Preferably, the M pieces of new input color information are color information of a division point when the N pieces of input color information in the data table are sequentially arranged and equally divided.

As a result, it is possible to easily obtain highly accurate total difference data. In order to achieve the above object, a data table calibration method for calibrating a data table for recording N pieces of input color information and N pieces of output color information respectively corresponding to the N pieces of input color information is provided. , M (M <N) pieces of new input color information and M pieces of new output color information corresponding to each of the M pieces of new input color information.
A reading step of reading M pieces of output color information corresponding to each of the pieces of new input color information; a step of reading each of the pieces of read out M pieces of output color information; A first calculation step of calculating a difference from each of them as M pieces of difference data, a second calculation step of calculating N pieces of difference data by interpolating the M pieces of difference data, A calibration step of calibrating the N pieces of output color information in the data table using the N pieces of difference data.

According to the present invention, it is possible to provide a data table calibrating method capable of calibrating a data table more easily.

[0023]

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing the overall configuration of a data calibration device according to an embodiment of the present invention. Referring to FIG. 1, the data proofreading apparatus uses RGB as input color information.
The lookup table data storage unit 101 stores the correspondence between the values and the CMY values as the output color information in a table, and a new correspondence is stored in a table to update the lookup table. An update table data storage unit 103, a difference data calculation unit 105 for calculating a difference between the lookup table data and the update table data, and a total difference data calculation unit 107 for calculating all the difference data from the difference data by the difference data calculation unit 105
And a new table data creation unit 109 for creating new lookup table data from the difference data and the lookup table data by the total difference data calculation unit 107
It is composed of

Look-up table data storage unit 101
Ideally, for each of the 8 bits of RGB as input color information, that is, for all 24 bits, CMY as output color information
It is desirable to store each 8-bit data in association with each other. However, in consideration of the storage capacity and the like, the size of each of the RGB is n bits (n <8).

Therefore, the input color information is RGB
This means that 2 ⁿ * 2 ⁿ * 2 ⁿ grid points in the color space have been selected. As output color information, there are also 2 ⁿ * 2 ⁿ * 2 ⁿ grid points in the CMY color space. Note that the grid points in this CMY color space are
Unlike the lattice points in the RGB color space, the relation between the lattice points is considered to be non-linear.

However, while the relationship between long-range gratings is non-linear, the relationship between short-range gratings can be regarded as approximately linear. Therefore, R
Output data corresponding to data between grid points in the GB color space can be obtained by performing linear interpolation from grid points in the CMY color space.

The update table data storage unit 103 updates the look-up table with a small number of data. Therefore, the update table data storage unit 103 is smaller than the size of each n bits of RGB as input color information in the lookup table storage unit 101.
Here, one-quarter of each of RGB, that is, each of RGB (n
-2) Bit size.

Therefore, the input color information is 2 ⁿ * 2 ⁿ * 2 ⁿ in the RGB color space of the look-up table.
2 ^(n-2) * 2 ^where each of the grid point data is reduced to 1/4
^(n-2) * 2 ^(n-2) grid points are selected. FIG. 2 schematically shows an example of grid point data in the RGB color space at this time. As shown in FIG. 2, grid points in the RGB color space are grid point data (RGB values) as input color information.

Since output color information (CMY values) is associated with each of the grid point data, the output color information is also 2 ^(n-2) * in the CMY color space.
2 ^{(n− 2)} * 2 ⁽ⁿ⁻²⁾ lattice points exist.

The relationship between the lattices in the CMY color space is also non-linear and not a simple relationship. In particular, CMY in lookup table data
Unlike the relation of the grid point data in the color space, the number of grid points is one quarter for each of CMY, so the relation between adjacent grid point data (the relation between short range grid points).
Are no longer simply considered to be linear here. Therefore, data between grid points cannot be obtained by simple linear interpolation as obtained by the lookup table data.

In order to explain this, FIG. 3 shows the relationship between the lookup table data and the update table data. Here, the one-dimensional original color space is selected on the horizontal axis, and the one-dimensional output color space is selected on the vertical axis.

Referring to FIG. 3, a point on solid curve C1 is
The old table data, that is, the grid point data stored in the lookup table data storage unit 101 in advance, and the points on the broken line curve C2 indicate the grid point data stored in the update table data storage unit 103. Is shown. Note that arrows indicate difference data between the two data.

The interval between adjacent grid point data in the update data table is about four times wider than the interval of the old table data, and if these are approximated to have a linear relationship, a curve C3 results. Therefore, this curve C
3 and the broken line curve C2 connecting the new table data are considerably different from each other, indicating that the linear approximation is too rough.

For this reason, adjacent grid point data in the update table data cannot be regarded as having an approximately linear relationship like the relationship between adjacent grid point data in the old data table, and the data between them cannot be considered. It can be seen that finding by linear interpolation is inappropriate.

Therefore, a data calculation procedure between update table data according to the embodiment of the present invention will be described below. FIG. 4 is a diagram showing a flow for creating new table data from old table data and updated table data. FIG. 5 is a diagram for explaining the flow of FIG. 5, the horizontal axis represents the one-dimensional original color space, the vertical axis represents the one-dimensional output color space, the points on the solid curve C1 represent the old table data, and FIG. Points on the broken line curve C2 indicate new table data.

Referring to FIG. 4, first, at step S401
, The old table data and the updated table data storage unit 1 in the lookup table data storage unit 101
03, the difference data is obtained from the update data. This corresponds to obtaining a solid arrow from a point on the solid curve C1 and a black point on the dashed curve C2 in FIG.

The number of old table data is n bits for each of RGB
2ⁿ* 2ⁿ* 2ⁿAnd the number of updated data is
2 for each quarter of the RGB data^(n-2)* 2^(n-2)
* 2 ^(n-2)Individual. Therefore, the number of difference data is also updated.
2 equal to the number of new data^(n-2)* 2^(n-2)* 2^(n-2)Individual
You.

Next, in step S403, the total difference data 2 ⁿ * 2 ⁿ from the 2 ⁽ⁿ⁻²⁾ * 2 ⁽ⁿ⁻²⁾ * 2 ⁽ⁿ⁻²⁾ difference data obtained in step S401. * 2 ⁿ pieces are required. That is, here, difference data is obtained for all the input data of the old data table. In FIG. 5, solid arrows and dotted arrows corresponding to all points on the solid curve C1 correspond.

Specifically, by simply performing linear interpolation between adjacent difference data, data therebetween, that is, interpolation difference data is obtained (dotted arrow in FIG. 5). As described above, the linear interpolation is performed because the relationship between adjacent grid point data in the update data table is non-linear and cannot be linearly interpolated, but the old table data and the new table data are not completely unrelated. Because it is possible. That is, the relationship between the difference data can be considered to be approximately linear.

Then, in step S405, the old text
Table data from the table data and all difference data.
Is created (in FIG. 5, all the points on the dashed curve C2)
Black point and white point). This is 2ⁿ* 2ⁿ* 2ⁿOld tabs
Bull data and 2ⁿ* 2 ⁿ* 2ⁿTotal difference data
It is determined by taking the sum (or difference) of did
Therefore, the number of new table data is equal to the number of old table data
New 2ⁿ* 2ⁿ* 2ⁿIndividual.

In this manner, the same number of new table data as the old table data can be predicted and obtained from the number of update data smaller than the number of old table data. Therefore, it is not necessary to obtain all the new table data from the beginning by experiment, and it is possible to calibrate the lookup table data more easily.

An experiment result when the lookup table is calibrated according to the data table calibrating method will be described below. First, a lookup table having a size of 9 * 9 * 9 for converting a color signal in the Lab color space into a CMY color signal for a printer was created. In order to create a table, a color chart is actually output using a printer, and color measurement of an output sample is performed. In addition,
The data between the lattice data is obtained by hexahedral interpolation using the nearest eight-point data.

Using this look-up table, color conversion processing was performed on 10 randomly selected input data, and printed out, the average color difference was ΔE <5.

Next, the color reproduction was changed by changing the setting conditions of the printer. The average color difference in this case was ΔE> 8.

Then, the look-up table was calibrated by the data table calibrating method in order to match the state of the printer in which the color reproduction was changing. As data for updating, 5 * 5 * 5 data is newly collected by the same experiment. The average color difference when printing was performed using the newly calibrated look-up table was ΔE <5, and returned to the initial state.

As described above, in comparison with the case where a data table is newly created from scratch, the calibration of the data table having the same color reproducibility as the initial state can be performed by using a very small number of update data. It becomes possible.

Therefore, the color reproduction can be easily corrected at the user's hand, which has been difficult in the past. In addition, the conditions for semi-automatic or automatic calibration are improved, and it becomes possible to incorporate the automatic calibration processing function in advance as needed. In this case, merits such as a reduction in the load on the device, the CPU, and the memory, and a reduction in the time required for automatic color measurement, which eliminates the user's waiting time, are also obtained.

The embodiments disclosed this time are illustrative in all respects, and should not be construed as limiting. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an overall configuration of a data calibration device according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating an example of grid point data in an RGB color space.

FIG. 3 is a diagram showing a relationship between lookup table data and update table data.

FIG. 4 is a diagram showing a flow for creating new table data from old table data and updated table data.

FIG. 5 is a diagram for explaining the flow of FIG. 4;

FIG. 6 is a flowchart illustrating a flow of a general color conversion process using a lookup table.

FIG. 7 is a diagram showing a typical configuration of a lookup table referred to.

[Explanation of symbols]

 Reference Signs List 101 Lookup table data storage unit 103 Update table data storage unit 105 Difference data calculation unit 107 Total difference data calculation unit 109 New table data creation unit

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Jun Mizuguchi 2-13-13 Azuchicho, Chuo-ku, Osaka City International Building Minolta Co., Ltd. F-term (reference) 5B057 CA01 CA08 CA12 CA16 CB01 CB08 CB12 CB16 CC01 CD06 CE18 CH07 CH08 5C077 LL17 MP08 PP32 PP33 PP68 PQ23 SS02 TT02 TT06 5C079 HB01 HB02 HB12 LA28 LB02 MA05 MA11 NA13 NA27 PA02 PA03

Claims (4)

[Claims] 1. A data table calibrating device for calibrating a data table for recording N pieces of input color information and N pieces of output color information corresponding to each of the N pieces of input color information, Acquiring means for acquiring M (M <N) new input color information and M new output color information corresponding to each of the M new input color information; Reading means for reading M pieces of output color information corresponding to each of the M pieces of new input color information; each of the read out pieces of M output color information; First calculating means for calculating a difference from each of the output color information as M difference data, and second calculating means for calculating N difference data by interpolating the M difference data And using the calculated N pieces of difference data, And a calibration means for calibration of the N output color information in the data table, the data table calibration device. 2. The data table calibrating device according to claim 1, wherein said second calculating means performs linear interpolation of said M difference data. 3. The method according to claim 2, wherein the M pieces of new input color information are color information of division points when the N pieces of input color information in the data table are arranged in order and equally divided. The data table calibrating apparatus according to claim 1 or 2. 4. A data table calibrating method for calibrating a data table for recording N pieces of input color information and N pieces of output color information respectively corresponding to the N pieces of input color information, wherein M An obtaining step of obtaining (M <N) pieces of new input color information and M pieces of new output color information corresponding to each of the M pieces of new input color information; A reading step of reading M output color information corresponding to each of the M new input color information; each of the read M output color information; and the obtained M new output A first calculation step of calculating a difference from each of the color information as M pieces of difference data, and a second calculation step of calculating N pieces of difference data by interpolating the M pieces of difference data. The calculated N difference data Using said and a calibration step of calibrating the N outputs color information, data table calibration method in the data table.