JP2023065885A - Image processing device, image processing method and program - Google Patents

Abstract

To perform a color conversion process efficiently using hardware resources.SOLUTION: Axis setting for deriving a point where a conversion target point ti and one of color components match is performed using a reference table. The output values of points Q1, Q2, Q3, Q4 between grid points where the conversion target point ti and the value in R axis direction are the same, are acquired from the reference table. Linear interpolation in a G axis direction is carried out using the acquired output values of Q1-Q4, and the output values of interpolation points R1, R2 between the points Q1-Q4 where the conversion target point ti and the value in G axis direction are the same are calculated. Linear interpolation in a B axis direction is carried out using the acquired output values of interpolation points R1, R2, and the output value of an interpolation point S between the interpolation points R1, R2 where the conversion target point ti and the value in B axis direction are the same, i.e., the conversion target point ti, is calculated.SELECTED DRAWING: Figure 6

Description

The present disclosure relates to image processing technology for color conversion processing.

2. Description of the Related Art Conventionally, a rendering method has been widely used in which outline information (also called edge information) of a figure is extracted from the coordinate information of the figure in order to draw the figure, and an image is formed based on the extracted outline information.

Documents to be printed that are provided as digital data (including characters, graphics, and photographic images) are often composed of color data. or When printing such a document on paper, etc., the document is converted to page description language (PDL) format data (PDL data) and sent to the printer, the PDL data is processed in the printer, and the image is transferred to the recording medium. Form on top. In order to print based on the PDL data, when RIP (Raster Image Processor) processes the PDL data, color conversion processing is executed by a CMS (color management system) or the like for the colors corresponding to each object in the PDL data. be.

This color conversion processing is desired to be speeded up, but in many cases, it is implemented by software, so the speed depends on the processing speed of the CPU that executes the software. Patent Document 1 proposes a method of using a color lookup table (color LUT) to speed up color conversion processing.

JP 2008-245029 A

However, with the method disclosed in Japanese Patent Laid-Open No. 2002-100000, in environments where the size of the cache memory is severely constrained, such as when a processor used in an embedded environment is used, the necessary color LUT cannot be developed all at once, and the color conversion process can be efficiently processed. There is a problem that it is difficult to

According to the present disclosure, an image processing device that converts an input value in an input color space made up of a plurality of color components into an output value in an output color space, wherein the lookup in which the input value and the output value are associated A plurality of outputs corresponding to a plurality of input values in the input color space in which an input value to be converted in the input color space and one of the plurality of color components have the same value by referring to a table. obtaining means for obtaining a value; and calculating means for performing an interpolation operation using a plurality of output values obtained by referring to the lookup table to calculate an output value corresponding to the input value to be converted. characterized by

In the present disclosure, it is possible to efficiently use hardware resources to perform color conversion processing.

Diagram showing a schematic configuration example of a general image forming unit FIG. 1 shows a hardware configuration example of an image processing apparatus according to Embodiment 1 of the present disclosure; Schematic diagram explaining internal processing tasks in the image processing apparatus according to Embodiment 1 of the present disclosure A diagram showing a configuration example of an image processing apparatus according to Embodiment 1 of the present disclosure Conceptual diagram explaining general interpolation processing A diagram showing an interpolation processing procedure according to the first embodiment of the present disclosure FIG. 11 is a diagram showing a configuration example of a printing system including an image processing apparatus according to Embodiment 2 of the present disclosure; Schematic diagram of analysis of color distribution for each scene in Embodiment 2 of the present disclosure FIG. 11 is a diagram showing a processing procedure for sharing color distributions of a plurality of scene-specific image data and synthesizing them into one according to the second embodiment of the present disclosure; FIG.

Hereinafter, embodiments of the present disclosure will be described in detail with reference to the accompanying drawings.

<Embodiment 1>
An overview of image processing in the printing system of this embodiment will be described. FIG. 1 shows a schematic configuration example of an image forming unit in the color printer 2010 of this embodiment. In the figure, 2200 to 2230 are color ink cartridges filled with yellow (Y) color ink, magenta (M) color ink, cyan (C) color ink, and black (K) color ink, respectively. Each of the ink cartridges 2200 to 2230 and the pressure pump 2160 are connected by independent pipes for supplying ink, and the ink of each color is pressure-fed from the pressure pump 2160 to the print head 2120 at a constant pressure. ing.

A recording sheet is supplied from the rear surface of the color printer 2010 and fixed by the platen 2110 and the front guide roller 2150 . The color printer 2010 is operated by pressing keys on the control panel 2180 , and the control board 2190 controls all operations of the color printer 2010 .

In order to print with the color printer 2010, a print control command and print data are sent to the control board 2190 via a network interface or the like (not shown). The network may be a wireless network, and is configured so that the printer is appropriately connected to an AP near the installation location. When the control board 2190 receives the print control command and the print data, for example, when RGB is specified in the color specification command, the internal color processing unit converts the RGB data into CMYK data, and then drives the print head 2120 . to print.

FIG. 2 is a diagram showing an example of the hardware configuration of an image processing apparatus that implements the control board 2190. As shown in FIG.

In FIG. 2, a CPU 220 executes programs such as an OS and general applications loaded from a nonvolatile storage device (hard disk, etc.) (not shown) to a RAM 224 such as a DRAM, and implements software functions described later. The RAM 224 functions as a main memory, work area, etc. of the CPU 220 .

Although the CPU 220 and the RAM 224 shown in FIG. 2 are connected by a bus, a cache memory is prepared therein and functions as a pre-stage PRE loading area for transferring the data of the RAM 224 to the CPU 220 . The cache implemented in the CPU 220 is often composed of two stages of high-speed SRAMs, one being the L1 cache 222 and the other being the L2 cache 223 . As methods of selecting internal lines for a cache, a number of schemes have been proposed for improving efficiency, such as full associative, direct map, and N-way set associative methods.

Note that the CPU 220 of the present embodiment is assumed to be a single-core CPU, but a multi-core CPU may be used, or a plurality of CPUs may be used.

In this embodiment, the following functions of the image processing apparatus are described as being realized by software, but each function may be implemented in the image processing apparatus by dedicated hardware.

FIG. 3 shows a schematic diagram explaining internal processing tasks in the image processing apparatus according to the first embodiment.

When the system task 311 detects that print job data (hereinafter referred to as PDL data) has been delivered, it issues a print start command to the PDL task 312 .

Upon receiving a print start command, the PDL task 312 passes the result of analyzing the drawing commands and print job setting information included in the PDL data to the image forming task 313 to start RIP (Raster Image Processor) processing.

When the image forming task 313 receives the analysis result of the PDL data, it starts forming image data (raster image data) while interpreting information such as graphics and characters described in the PDL data, and generates the generated raster image data. to control task 314 . In the image forming task 313, when an image is embedded in the transferred PDL data, processing is performed to read the corresponding image data from the auxiliary storage device as necessary. The speed at which the image data is read into the memory at this time is, for example, about 10 MB/sec in a typical system. On the other hand, the raster image data generated by the image forming task 313 is sequentially generated in band units, for example, and passed to the PDL task 312 as image data. The average data generation speed assumed by the system is assumed here to be about 6 MB/sec.

Upon receiving the raster image data, the control task 314 causes the printer engine to form an image based on the received raster image data. A control signal from the control task 314 is configured to be supplied timely according to a communication interface (CI) (not shown) and a defined protocol. The speed at which data is received by the printer engine varies depending on the type of printer engine, but here, for example, about 1 MB/sec is assumed as a typical example. In any case, the image formation/transfer band and processing time are set, and the modules are designed so that bottlenecks do not occur in each processing unit.

FIG. 4 shows a configuration example of each module of the image processing apparatus according to this embodiment. FIG. 4 illustrates three modules: user system 411 , CMS module 412 , and RIP module 413 .

A user system 411 is a module that executes the above-described system task 311 and PDL task 312, and is a module that controls the entire print process. When a job is input, it is passed to the lower RIP module 413 to generate raster image data for print output.

A CMS module 412 is a module in charge of CMS (color management system) processing, and is configured to be called from the RIP module 413 for color conversion processing. The CMS module 412 stores setting information for color conversion, a color conversion converter (CONV-A, etc.) appropriately generated based on the setting information, and a table that holds decompressed data used in part of the conversion process. have departments, etc. The converter for each color conversion performs interpolation calculation processing from data such as ICC profile and color LUT, and calculates the desired conversion value. can. By holding a certain amount of input values and output values once converted in the cache memory, arithmetic processing can be omitted when the same color is repeatedly converted. In addition, fixed values such as .gamma. values can be stored in the expanded data, and conversion can be performed only by referring to the values without calculating each time.

Next, interpolation processing in color conversion processing will be described with reference to FIG. FIG. 5 schematically shows the input values of the color LUT in three dimensions, and assumes that the input color space is the three-dimensional color space of the RGB color space. Note that the input color space is not limited to the RGB color space, and may be a CMYK color space having a different number of dimensions. Any color space can be used for the input color space and the converted output color space.

This three-dimensional color space is partitioned by a plurality of unit cells, for example unit cells that divide each axis of the input color space into 15 divisions. When the gradation of each color is represented by 8 bits, the input values take values from 0 to 255, and in this case, each axis of each unit cell has 17 input values. The output value corresponding to the conversion target point ti is obtained by first deriving the output value corresponding to the grid point (grid point) of the unit cell containing the conversion target point ti from the color LUT, and using the output value of the derived grid point. It is obtained by performing interpolation processing. Conventionally, linear interpolation is performed for interpolation processing between grid points. As linear interpolation, 8-point interpolation, 5-point interpolation, 4-point interpolation, etc. are often used, and in any case, linear calculation is performed starting from the value of each grid point. Since addition and accumulation processing is executed in this calculation, 5-point interpolation or 4-point interpolation processing with a small number of points is advantageous in terms of speed, but in consideration of interpolation accuracy, 8-point interpolation is performed in this embodiment. A case will be described.

In the case of 8-point interpolation, first, a point located between grid points having the same value as the conversion target point ti is derived with respect to one predetermined color component, for example, the R-axis direction. In the example shown in FIG. 5, point Q1 is derived from grid points P1 and P2, point Q2 is derived from grid points P3 and P4, point Q3 is derived from grid points P5 and P6, and point Q4 is derived from grid points P7 and P8. Next, using the points Q1 to Q4, points R1 and R2 positioned between grid points having the same value as the conversion target point ti are derived with respect to color components other than the R component, for example, the G axis direction. Finally, the points R1 and R2 are used to derive the remaining B component, that is, the point S located between the grid points having the same value as the conversion target point ti in the B-axis direction, that is, the conversion target point ti.

In this embodiment, only the first points Q1, Q2, Q3, and Q4 in the interpolation process are derived, instead of being calculated by interpolation calculations, precalculated values recorded in a reference table consisting of a color LUT are obtained. (Fig. 6(a)).

FIG. 6B shows a flowchart of interpolation processing according to this embodiment. In this embodiment, first, in order to derive a point where one of the color components matches the conversion target point ti, all input values on a line segment parallel to the R axis and passing through the grid points and the corresponding output values is stored in association with the color LUT as a reference table.

In S611, the reference table is used to set an axis (SET_TABLE_ON function) for deriving a point where one of the conversion target points ti and one of the color components match. Then, a reference table consisting of a color LUT in which input values on a line segment parallel to the set axis and passing through the grid points and corresponding calculated output values are stored in association with each other is loaded onto the memory.

In S612, output values of points Q1, Q2, Q3, and Q4 between grid points having the same value in the R-axis direction as the conversion target point ti are acquired from the reference table. Here, as arguments of the function FUNC1 that acquires the output value from the reference table, the origin information, which is the input value of the grid point adjacent to the conversion target point ti, and the difference value in the R-axis direction between the grid point and the conversion target point ti FR is used.

In S613, the output values of the points Q1 to Q4 acquired in S612 are used to perform linear interpolation calculation in the G-axis direction, and the interpolation points between the points Q1 to Q4 having the same value in the G-axis direction as the conversion target point ti are obtained. R1 and R2 are determined and their output values are calculated.

In S614, the output values of the interpolation points R1 and R2 obtained in S613 are used to perform a linear interpolation calculation in the B-axis direction, and the interpolation points R1 and R2 having the same value in the B-axis direction as the conversion target point ti are calculated. An output value of the interpolation point S, that is, the conversion target point ti is calculated.

In this embodiment, the output value is obtained from the reference table in S612 and the interpolation calculation is omitted, thereby reducing the number of multiplications in the interpolation calculation. Therefore, it is possible to improve the interpolation calculation speed especially when the number of colors is large. In addition, the reference table used is not a color LUT that corresponds to all input values, but is limited to a color LUT that corresponds to input values on a line segment that is parallel to the set axis and passes through the grid points, thus reducing memory usage. can do For these reasons, even in an environment with a small cache size such as an embedded processor, the memory capacity is less likely to become tight, so it is possible to make the most of the CPU's performance.

<Embodiment 2>
FIG. 7 shows a schematic configuration of a color printing system for printing based on PDL data including an image captured by a digital single-lens camera or the like according to the second embodiment. In the figure, a host PC 710 and a printer controller 720 are wired or wirelessly connected via a communication interface (CI) (not shown).

In this system, since the bandwidth for data transfer on the transmission line is not large, the data size is made compact as PDL data, and the print job is sent. Data is transferred to the printer controller 720 side. The printer controller 720 side passes the received PDL data to the RIP module 722 via the print control unit 721 . After receiving the PDL data, the RIP module 723 is configured to appropriately perform color matching (color conversion processing) using the CMS module 722, generate raster image data, and deliver the raster image data to the print control unit 721. .

In the first embodiment, the color LUT corresponding to the input values on the line segment parallel to the set axis and passing through the grid points in the color space is developed in the memory as a reference table. It may be larger than that. Therefore, in the second embodiment, the bias in color distribution is checked in advance for each image, and the color LUT data of unused colors is configured so as not to be developed in the memory.

Considering the case of a natural image captured by a digital camera or the like, skin-colored color gamuts frequently appear in captured images in which a person is the subject. In that case, the blue color gamut of the sky and the sea will appear frequently. In this way, it is possible to grasp in advance the pattern of the color distribution of the captured image by examining captured images for various shooting scenes and combining them.

FIG. 8A shows a color distribution analysis chart for each shooting scene. Looking at these from a bird's eye view, there are color gamuts with high appearance frequency and color gamuts with low appearance frequency in each shooting scene, and the color distribution is biased. Therefore, in this embodiment, unlike the first embodiment, the color LUT corresponding to all areas on line segments parallel to the set axis and passing through the grid points is not used. The configuration is such that only the LUT is expanded on the memory as a reference table. Here, the color space is divided into 9.times.9.times.9 unit grids, and the "high frequency appearance color distribution by scene" is specified by determining the frequency of appearance for each color gamut corresponding to each unit grid. Only the color LUT of the color gamut corresponding to the unit grid determined to have a high appearance frequency is developed on the memory as a reference table.

Further, FIG. 8B shows a "common high-frequency color distribution" representing a color gamut with a high appearance frequency in the color distribution obtained by superimposing the color distributions of the shooting scenes. By using a reference table generated based on this common high frequency color distribution, it is possible to efficiently omit the color LUT for color gamuts with low frequency of appearance regardless of the shooting scene.

FIG. 9A is a schematic diagram showing the process of generating the common high-frequency color distribution shown in FIG. 8, and FIG. shows the processing steps for generating The processing for generating this common high-frequency color distribution is performed in advance using an information processing device, and the generated common high-frequency color distribution data is sent to the printer controller 720 shown in FIG. 7 together with the color LUT and the like. is stored in advance in an auxiliary storage device or the like.

As a method of synthesizing the color distribution of image data for each shooting scene, as shown in FIG. Assume that the color distribution of the next scene is synthesized with the color distribution (composite color distribution). For example, in the example shown in FIG. 8B, first, the color distribution of the "fireworks" scene and the color distribution of the "pet" scene are combined to generate a combined color distribution. The color distribution of the "snow scene" scene is further synthesized. Note that the color distribution synthesis method is not limited to this, and other methods may be used as long as they can be appropriately synthesized.

In S911, a weight parameter is set as an initialization parameter. The weight parameters are prepared for each grid. For example, 729 weight parameters are set when using a unit grid divided equally into 9 in the axial direction, and 3375 parameters are set when using a unit grid divided into 15 divisions. These weight parameters are usually set to 1 as an initial value, but the user may set any value.

In S912, one image data for synthesizing the color distribution is selected from a plurality of prepared image data for each shooting scene.

In S913, one unit grid for synthesizing the color distribution is selected from a plurality of unit grids for dividing the color distribution. For example, when there are 9×9×9 unit cells, unselected unit cells are selected in order from the starting unit cell.

In S914, the color distribution of the unit cell selected in S913 is combined with the combined color distribution and the color distribution of the image data of the shooting scene selected in S912, and the number of colors distributed within the unit cell is counted.

In S915, whether the value obtained by multiplying the number of colors distributed within the unit cell by the weight parameter exceeds a predetermined threshold value (for example, a number corresponding to 50% of the total number of possible colors within each unit cell). determine whether or not If the number of colors distributed within the unit grid exceeds the threshold, the process proceeds to S916, and if the number of colors distributed within the unit grid does not exceed the threshold, the process proceeds to S917.

In S916, when the number of colors distributed within the unit cell exceeds the threshold value, processing such as raising a flag as a color gamut to be included in the reference table developed in the memory is performed.

In S917, it is determined whether or not there are no unselected unit cells other than the currently selected unit cell and it is the last unit cell. If there is no grid, the process proceeds to S918.

In S918, it is determined whether or not there is image data of other unselected imaging scenes in addition to the currently selected image data. If there is no such image data, the series of processing ends.

The common high-frequency color distribution thus obtained is referred to when the CMS module 722 or the like develops the color LUT in the cache memory, and only the color LUT of the color gamut corresponding to the flagged unit grid is stored in the cache memory. allow it to expand to

In the process shown in FIG. 9A, the generated common scene is inspected as necessary for any problem, and based on the inspection result, it is determined whether or not the weight parameter and the effective area are determined. It is also possible to provide a mechanism for appropriately adjusting the threshold to be used and reconfiguring it. Also, although this part includes a trial and error part, for example, a mechanism for automatic adjustment using a learning function or the like may be introduced.

(Other examples)
The present invention supplies a program that implements one or more functions of the above-described embodiments to a system or apparatus via a network or a storage medium, and one or more processors in the computer of the system or apparatus reads and executes the program. It can also be realized by processing to It can also be implemented by a circuit (for example, ASIC) that implements one or more functions.

411 User System 412 CMS Module 413 RIP Module

Claims (10)

An image processing device that converts an input value in an input color space consisting of a plurality of color components into an output value in an output color space,
By referring to a lookup table in which the input value and the output value are linked, the input value to be converted in the input color space and one color component of the plurality of color components are the same value. obtaining means for obtaining a plurality of output values corresponding to a plurality of input values in the input color space;
calculating means for performing an interpolation operation using a plurality of output values obtained by referring to the lookup table to calculate an output value corresponding to the input value to be converted;
An image processing device comprising: The lookup table includes: a plurality of input values on a line segment passing through lattice points of a plurality of unit lattices dividing the input color space and parallel to the axis of the one color component; It holds the corresponding multiple output values in association with each other,
2. The image processing apparatus according to claim 1, wherein: The calculation means calculates an input in the input color space that has one more color component of the same value as the input value to be converted than the plurality of input values corresponding to the plurality of output values obtained by referring to the lookup table. determining a value, and calculating an output value corresponding to the input value by performing an interpolation operation using the output value obtained by referring to the lookup table;
3. The image processing apparatus according to claim 1, comprising: The calculating means determines an input value in the input color space that has one more color component of the same value as the input value to be converted than the input value corresponding to the output value calculated by the calculating means, and determines the input value Calculate an output value corresponding to by performing an interpolation operation using the calculated output value,
4. The image processing apparatus according to claim 3, characterized by: the lookup table includes only a color LUT of a predetermined color gamut in which the input values to be converted appear frequently;
5. The image processing apparatus according to any one of claims 1 to 4, characterized by: The color LUT of the predetermined color gamut included in the lookup table is a color LUT of a color gamut that frequently appears in a plurality of different image data.
6. The image processing apparatus according to claim 5, characterized by: the number of dimensions of the input color space and the output color space is 3 or 4, respectively;
7. The image processing apparatus according to any one of claims 1 to 6, characterized by: one of the input color space and the output color space is an RGB color space and the other is a CMYK color space;
8. The image processing apparatus according to any one of claims 1 to 7, characterized by: An image processing method for converting input values in an input color space consisting of a plurality of color components into output values in an output color space,
By referring to a lookup table in which the input value and the output value are linked, the input value to be converted in the input color space and one color component of the plurality of color components are the same value. obtaining a plurality of output values corresponding to a plurality of input values in the input color space;
performing an interpolation operation using a plurality of output values obtained by referring to the lookup table, and calculating an output value corresponding to the input value to be converted;
An image processing method characterized by comprising: A program for causing a computer to function as the image processing apparatus according to any one of claims 1 to 8.

Images