JP2014143512A - Image processing device and image processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processing device which can improve image quality with a single run of filter processing executed based on an optimum filter coefficient which can be calculated with a small amount of computation even if variable magnifications used in magnification processing performed on image data differ in the longitudinal direction and the lateral direction.SOLUTION: An image processing device includes: first and second conversion matrix setting sections 301 and 302 for setting first conversion matrices in accordance with variable magnifications for a longitudinal direction and a lateral direction; a selecting section 303 for selecting a reference filter coefficient set in response to (i) a prescribed image processing condition different from the variable magnifications in the longitudinal and lateral directions or an image processing condition which adds the variable magnifications for the longitudinal and lateral directions to the prescribed image processing condition and (ii) one or more filter coefficient sets which are prestored; coefficient conversion sections 304a and 304b for primarily converting a matrix indicated by the selected reference filter coefficient set by using the first conversion matrix and a second conversion matrix; and a filter processing section 26 for performing filter processing on image data by using the converted filter coefficient set.

Description

  The present invention relates to an image processing apparatus and an image processing method, and more particularly, to an image processing apparatus and an image processing method that improve the image quality of an output image based on image data by performing filter processing on the image data.

  2. Description of the Related Art Conventionally, scanners that optically read originals to acquire image data, and image forming apparatuses such as copiers and multifunction peripherals equipped with scanners have been distributed. In these apparatuses, when image data (image data) obtained by reading a document is output or printed, a filter process may be performed as one method for improving the image quality of the output image. In addition, there is an apparatus having a configuration in which a document reading speed (scanning speed) can be set fast in order to increase the processing speed.

  When the scanning speed is set high, the reading resolution in the main scanning direction (direction perpendicular to the scanning direction of the scanner) of the document may differ from the reading resolution in the sub-scanning direction (direction parallel to the scanning direction of the scanner). . In such a case, when outputting image data obtained by reading a document, it is necessary to use different scaling factors for the scaling process in the main scanning direction and the scaling process in the sub-scanning direction. In addition, there is an apparatus having a configuration capable of performing a scaling process at different scaling ratios in the main scanning direction and the sub-scanning direction of image data depending on settings from the user.

  When the scaling process is performed at different scaling ratios in the main scanning direction and the sub-scanning direction of the image data, the spatial frequency characteristics of the image data in the main scanning direction and the sub-scanning direction are different. Therefore, the image quality of the output image may be reduced due to the occurrence of moire or the like.

  In order to avoid this, it is necessary to prepare filter types for each scaling factor. However, the combination of scaling factors for main scanning and sub-scanning is very large, and as many filters are required for that combination. As a result, the memory capacity required to hold the data increases, and it takes time to check and manage the designed filter.

  In view of this, Japanese Patent Laid-Open No. 2004-228688 proposes a technique for performing filter processing in consideration of a scaling factor in scaling processing performed in the main scanning direction and a scaling factor in scaling processing performed in the sub-scanning direction. Specifically, according to the magnification (magnification) specified in the main scanning direction, the apparatus described in Patent Document 1 applies to the image data before or after the magnification process as the magnification is higher. And strong MTF (modulation transfer function) correction (enhancement filter correction). Thereafter, according to the magnification in the sub-scanning direction specified separately from the magnification in the main scanning direction, the higher the magnification, the stronger MTF correction is performed on the image data that has been MTF corrected in the main scanning direction. Thus, appropriate MTF correction is performed on the image data before or after the scaling process.

  Patent Document 2 proposes a technique for switching a parameter for correcting a correction level in the main scanning direction and sub-scanning direction of an image, specifically, a characteristic of a filter used for spatial filter processing in accordance with a change in image magnification. Has been. As a result, image deterioration due to a change in image magnification is prevented.

  Further, in Patent Document 3, when the scaling factor is different between the main scanning direction and the sub-scanning direction, filter coefficients corresponding to each scaling factor are selected from a plurality of preset filter coefficients, and then the two filters are selected. There has been proposed a technique for generating an appropriate filter coefficient by converting coefficients into a spatial frequency component, combining them, and performing inverse conversion. That is, in this technique, only the filter coefficient corresponding to the standard variable magnification is set in advance, and when other magnifications are used, calculation is performed with the spatial frequency component, and a filter optimum for the magnification is generated.

Japanese Patent No. 2789560 JP-A-2-29072 JP 2009-267835 A

  However, the technique disclosed in Patent Document 1 has a problem that the processing time becomes long because MTF correction is performed twice. In the technique disclosed in Patent Document 2, a filter coefficient set optimized in accordance with a scaling factor in the main scanning direction (vertical direction) and a scaling factor in the sub-scanning direction (horizontal direction) is stored in a memory in advance. Therefore, there is a problem in that the memory capacity required for this purpose increases.

  The technique disclosed in Patent Document 3 is for solving the above problems, but the following problems remain. First, since the coefficient calculated by the inverse transformation is generally an irrational number, a processing step for rationalizing the filter coefficient is actually required. For rationalization, the total numerator after dividing all coefficients is generally set to be almost equal to the denominator (hereinafter called the division coefficient). Since there are two filter coefficients serving as the reference, the process of determining which filter coefficient division coefficient is selected or how to synthesize the division coefficients of the two filter coefficients is also required. Furthermore, in an actual device, the settable numerical values are limited to very limited integers, and therefore, non-optimal filter coefficients are generated due to errors or bugs that may occur when returning from spatial frequency components to filter coefficients. There is also a possibility.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a scaling factor for scaling processing performed in the vertical direction of image data and a scaling factor for scaling processing performed in the horizontal direction. Is to provide an image processing apparatus and an image processing method capable of realizing an improvement in image quality by a single filtering process based on an optimum filter coefficient that can be calculated with a small amount of calculation.

  In order to solve the above-described problem, a first technical means of the present invention is an image processing apparatus that performs a filtering process on image data for forming an image in which a plurality of pixels are two-dimensionally arranged. A first transformation matrix setting unit for setting a first transformation matrix according to a vertical scaling factor when the scaling process is performed on the image data, and a scaling process performed on the image data. A second transformation matrix setting unit for setting a second transformation matrix according to a scaling factor in the horizontal direction, and (i) predetermined image processing conditions different from the scaling factor in the vertical direction and the scaling factor in the horizontal direction, Alternatively, a reference filter based on the image processing condition obtained by adding the vertical scaling factor and the horizontal scaling factor to the predetermined image processing condition, and (ii) one or more filter coefficient sets stored in advance Selection section for selecting coefficient set and selection by the selection section A coefficient conversion unit that performs primary conversion on the matrix indicated by the reference filter coefficient set using the first conversion matrix and the second conversion matrix set by the first conversion matrix setting unit and the second conversion matrix setting unit And a filter processing unit that performs a filter process on the image data using the filter coefficient set after the primary conversion by the coefficient conversion unit.

  According to a second technical means of the present invention, in the first technical means, the size of the filter coefficient set after the primary conversion by the coefficient conversion unit can be different from the size of the reference filter coefficient set. The image processing apparatus further includes a size adjustment unit that adjusts the filter coefficient set after the primary conversion to match the size of the reference filter coefficient set and / or a predetermined size, and the filter processing unit is adjusted by the size adjustment unit The filter processing is performed using the filter coefficient set.

  According to a third technical means of the present invention, in the first or second technical means, the first transformation matrix is only a rational number generated by a predetermined calculation using one variable based on the scaling factor in the vertical direction. The second transformation matrix is a triangular matrix consisting only of rational numbers generated by a predetermined calculation using one variable based on the scaling factor in the horizontal direction. is there.

  According to a fourth technical means of the present invention, in any one of the first to third technical means, the vertical direction is changed according to the condition included in the predetermined image processing condition or the predetermined image processing condition. The conditions included in the image processing conditions including the magnification and the horizontal scaling factor are each divided into a plurality of sections, and in each section, the first conversion matrix, the second conversion matrix, and / or the reference filter This is characterized by switching coefficient sets.

  According to a fifth technical means of the present invention, there is provided an image processing method for performing filter processing on image data for forming an image in which a plurality of pixels are two-dimensionally arranged, wherein the first transformation matrix setting unit includes the image A first setting step for setting a first transformation matrix in accordance with a scaling factor in the vertical direction when the scaling process is performed on the data, and a second transformation matrix setting unit for scaling the image data A second setting step for setting a second transformation matrix according to a scaling factor in the horizontal direction when the processing is performed; and a selection unit: (i) What are the vertical scaling factor and the horizontal scaling factor? Different predetermined image processing conditions, or image processing conditions obtained by adding the vertical scaling factor and the horizontal scaling factor to the predetermined image processing conditions, and (ii) one or more filter coefficient sets stored in advance And select a reference filter coefficient set based on A matrix indicated by the reference filter coefficient set selected in the selection step, and the first conversion matrix and the second conversion matrix set in the first setting step and the second setting step. A conversion step for performing a primary conversion using a conversion matrix; and a filter processing unit for performing a filter process on the image data using the filter coefficient set after the primary conversion in the conversion step. It is characterized by.

  According to the present invention, even when the scaling factor of the scaling process performed in the vertical direction of the image data is different from the scaling factor of the scaling process performed in the horizontal direction, the calculation can be performed with a small amount of calculation. Image quality can be improved by a single filtering process based on the optimum filter coefficient.

1 is a block diagram illustrating a configuration example of an image forming apparatus including an image processing apparatus according to an embodiment of the present invention. FIG. 2 is a diagram illustrating a general configuration of a filter used in a spatial filter processing unit in the image forming apparatus of FIG. 1. FIG. 2 is a block diagram illustrating a configuration example of a filter coefficient setting unit in the image forming apparatus of FIG. 1. FIG. 10 is a block diagram illustrating another configuration example of the filter coefficient setting unit in the image forming apparatus of FIG. 1. FIG. 5 is a flowchart for explaining an example of a filter size reduction process in a size correction unit in the filter coefficient setting unit in FIG. 4. It is a figure which shows an example of the filter used for the window function process in the window function process part in the filter coefficient setting part of FIG. FIG. 6 is a diagram for explaining an example of filter coefficient setting processing in a filter coefficient setting unit in the image forming apparatus of FIG. 1 and is a diagram illustrating an example of a reference filter coefficient set. FIG. 6 is a diagram for explaining an example of filter coefficient setting processing in a filter coefficient setting unit in the image forming apparatus of FIG. 1, and is a diagram illustrating an example of a transformation matrix. FIG. 10 is a diagram for explaining an example of filter coefficient setting processing in a filter coefficient setting unit in the image forming apparatus of FIG. 1, and is a diagram illustrating another example of a transformation matrix. FIG. 10 is a diagram for explaining an example of filter coefficient setting processing in a filter coefficient setting unit in the image forming apparatus of FIG. 1, and is a diagram illustrating another example of a transformation matrix. It is a figure which shows the filter coefficient set of the result of having converted the reference | standard filter coefficient set of FIG. 7 with the conversion matrix of FIG. 8A. It is a figure which shows the filter coefficient set of the result of having converted the reference | standard filter coefficient set of FIG. 7 with the conversion matrix of FIG. 8B. It is a figure which shows the filter coefficient set of the result of having converted the reference | standard filter coefficient set of FIG. 7 with the conversion matrix of FIG. 8C. It is a figure which shows the frequency characteristic of the main scanning direction of the filter coefficient set of FIG. 9A-FIG. 9C. FIG. 7 is a flowchart for explaining an example of filter coefficient setting processing in a filter coefficient setting unit in the image forming apparatus of FIG. 1. FIG. 12 is a flowchart following FIG. 11.

Hereinafter, an image processing apparatus according to an embodiment of the present invention will be described by taking an image forming apparatus including the image processing apparatus as an example.
FIG. 1 is a block diagram illustrating a configuration example of an image forming apparatus including an image processing apparatus according to an embodiment of the present invention. An image forming apparatus 100 illustrated in FIG. 1 is, for example, a digital color copying machine, a color image input apparatus 1, a color image processing apparatus 2 (image processing apparatus), a color image output apparatus 3 as an image forming unit, and various operations. An operation panel 4 for performing the above is provided. Image data of analog signals of RGB (R: red, G: green, B: blue) obtained by reading a document with the color image input device 1 is output to the color image processing device 2, and the color image processing device 2 Then, a predetermined process is performed, and a digital color signal of CMYK (C: cyan, M: magenta, Y: yellow, K: black) is output to the color image output device 3.

  The color image input device 1 is a scanner having, for example, a CCD (Charged Coupled Device), reads a reflected light image from a document image in which a plurality of pixels are arranged two-dimensionally as an RGB analog signal, and reads the read RGB signal Is output to the color image processing apparatus 2. The color image output device 3 is a printer such as an electrophotographic system or an inkjet system that outputs an image in which a plurality of pixels are two-dimensionally arranged on a recording sheet based on image data of an original image. The color image output device 3 may be a display device such as a display.

  The color image processing apparatus 2 is an example of an image processing apparatus according to the present invention, and includes an A / D conversion unit 20, a shading correction unit 21, an input tone correction unit 22, a region separation processing unit 23, a color correction unit 24, and a black color. CPU for controlling the operation of the generated under color removal unit 25, the spatial filter processing unit 26, the scaling unit 27, the output tone correction unit 28, the tone reproduction processing unit 29, the filter coefficient setting unit 30, and these hardware units (Central Processing Unit) or DSP (Digital Signal Processor) or the like, and is configured by an ASIC (Application Specific Integrated Circuit) or the like.

  The A / D (analog / digital) converter 20 converts the RGB signal input from the color image input device 1 into, for example, a 10-bit digital signal, and outputs the converted RGB signal to the shading correction unit 21. .

  The shading correction unit 21 performs a correction process for removing various distortions generated in the illumination system, the imaging system, the imaging system, and the like of the color image input apparatus 1 on the input RGB signal, and the corrected RGB signal is converted into a corrected RGB signal. Output to the input tone correction unit 22.

  The input tone correction unit 22 is an image processing system employed in the color image processing apparatus 2 such as a density signal for the RGB signal (RGB reflectance signal) from which various distortions have been removed by the shading correction unit 21. In addition, the image signal is converted into an easy-to-handle signal, image quality adjustment processing such as color balance adjustment, background density removal or contrast is performed, and the processed RGB signal is output to the region separation processing unit 23.

  Based on the input RGB signal, the region separation processing unit 23 determines whether each pixel in the input image belongs to, for example, a character region, a dot region, a photographic region, or other regions, Separate the pixels. Based on the separation result, the region separation processing unit 23 outputs a region identification signal indicating to which region each pixel belongs to a color correction unit 24, a black generation and under color removal unit 25, a spatial filter processing unit 26, and a floor. Output to the key reproduction processing unit 29. The region separation processing unit 23 outputs the input RGB signal as it is to the subsequent color correction unit 24.

  The color correction unit 24 converts the RGB signal input from the region separation processing unit 23 into a CMY color space according to each region indicated by the region identification signal input from the region separation processing unit 23, and outputs a color image output device. The color correction is performed in accordance with the characteristic 3 and the corrected CMY signal is output to the black generation and under color removal unit 25. Specifically, the color correction unit 24 performs a process of removing color turbidity based on the spectral characteristics of the CMY color material including unnecessary absorption components in order to achieve faithful color reproduction.

  The black generation and under color removal unit 25 generates a K (black) signal based on the CMY signal input from the color correction unit 24 in accordance with each region indicated by the region identification signal input from the region separation processing unit 23. In addition, a black generation process is performed, and under color removal processing is performed to generate a new CMY signal by subtracting the K signal from the input CMY signal, and the generated CMYK signal is output to the spatial filter processing unit 26.

  An example of black generation processing in the black generation and under color removal unit 25 will be described. For example, in the case of general processing such as black generation processing using skeleton black, the input / output characteristics of the skeleton curve are y = f (x), the input data are C, M, and Y, and the output data is C ′. , M ′, Y ′, and K ′, and the UCR (Under Color Removal) rate is α (0 <α <1), the data output by the black generation and under color removal processing is K ′ = f { min (C, M, Y)}, C ′ = C−αK ′, M ′ = M−αK ′, and Y ′ = Y−αK ′.

  The color image processing apparatus 2 according to the present invention is an apparatus that performs filter processing (filter processing with computation using filter coefficients) on image data for forming an image in which a plurality of pixels are two-dimensionally arranged. For this purpose, a spatial filter processing unit 26 is provided.

  The spatial filter processing unit 26 includes each region indicated by the region identification signal input from the region separation processing unit 23 and a filter coefficient setting unit for the CMYK signal that has been subjected to black generation processing by the black generation and under color removal unit 25. Spatial filter processing using a digital filter (filter coefficient set) based on the filter coefficient set set by 30 is performed, and the processed image data (CMYK signal) is output to the scaling unit 27. In the spatial filter process, filtering is performed by multiplying the pixel of interest by the filter coefficient corresponding to the spatial position for each of the pixel values of the pixel of interest and neighboring pixels. Thereby, the spatial frequency characteristics of the image data are corrected, and blurring and graininess deterioration of the output image in the color image output device 3 are prevented.

  More specifically, the spatial filter processing unit 26 performs, for each region indicated by the region identification signal, the region identification signal (that is, the type of the region) by the filter coefficient setting unit 30 for each region indicated by the region identification signal. The spatial filter processing is performed using the filter coefficient set set corresponding to (). For example, the spatial filter processing unit 26 performs sharp enhancement processing on the regions separated into character regions by the region separation processing unit 23 to enhance the reproducibility of black characters or color characters, and emphasizes high frequency components. In addition, the spatial filter processing unit 26 performs a low-pass filter process for removing the input halftone component on the region separated by the halftone dot region in the region separation processing unit 23.

  In the following, for simplification of description, the case where the filter coefficient setting unit 30 and the spatial filter processing unit 26 have one type of region will be described. There is a reference filter coefficient set for each type of region separated by the processing unit 23, and the filter coefficient set calculated as a result and used by the spatial filter processing unit 26 is also different for each type of region identification signal. . Therefore, for example, if the region separation processing unit 23 separates the region into two types, that is, a character region and a halftone dot region, the reference filter coefficient set and the filter coefficient set used are for the character region and the halftone dot region. There will be two types.

  The scaling unit 27 enlarges or reduces the CMYK signal input from the spatial filter processing unit 26 based on the scaling factors in the main scanning direction and the sub-scanning direction specified by the user via the operation panel 4. (Variation processing) is performed and output to the output tone correction unit 28. The scaling process performed on the image data (CMYK signal) is performed by interpolating pixel data (in the case of increasing the resolution) or thinning out (in the case of decreasing the resolution) in the main scanning direction and the sub-scanning direction. It is realized by doing. Further, the scaling process may be optically performed in the sub-scanning direction, and the scaling process by interpolation may be performed in the main scanning direction.

  In the present embodiment, the scaling factor is not limited to the scaling factor input (designated) by the user via the operation panel 4 when the image data is enlarged or reduced and output. For example, when the original image is read by changing the reading speed by the color image input device 1 (faster or slower), converted to the same size image data (image magnification is 100%), and output to the color image output device 3 In addition, the scaling factor in the interpolation processing performed on the image data in the sub-scanning direction is also included.

  For example, when the standard reading resolution (reading resolution) by the color image input device 1 is 600 dpi (dot per inch), that is, 600 dpi × 600 dpi, when the reading speed by the color image input device 1 is doubled, for example, For example, the reading resolution is 600 dpi × 300 dpi, and the read image data is reduced to 50%. When the reduced image data is converted into equal-size image data and output to the color image output device 3, the scaling unit 27 performs conversion processing (interpolation processing) into equal-size (100%) image data. Done.

  As the interpolation process, a nearest neighbor method, a bilinear method, a bicubic method, or the like can be used. The nearest neighbor method is a method in which a pixel value of an existing pixel closest to a pixel to be interpolated (interpolated pixel) or a pixel value of an existing pixel having a predetermined positional relationship with respect to the interpolated pixel is used as the pixel value of the interpolated pixel. is there. The bilinear method is a method in which the pixel values of four existing pixels surrounding the interpolation pixel are weighted with values proportional to the distance from the interpolation pixel, and the average value of the obtained values is used as the pixel value of the interpolation pixel. In the bicubic method, a value calculated based on pixel values of a total of 16 existing pixels obtained by adding 12 existing pixels surrounding them to 4 existing pixels surrounding the interpolation pixel is used as the pixel value of the interpolation pixel. Is the method.

  The output tone correction unit 28 performs output tone correction processing for converting the CMYK signal (density signal) input from the scaling unit 27 into a halftone dot area ratio that is a characteristic value of the color image output device 3. The CMYK signal after the output tone correction processing is output to the tone reproduction processing unit 29.

  The gradation reproduction processing unit 29 performs predetermined processing on the CMYK signal input from the output gradation correction unit 28 based on each region indicated by the region identification signal input from the region separation processing unit 23. For example, the gradation reproduction processing unit 29 uses a high-resolution screen suitable for reproducing high-frequency components in the color image output device 3 in order to improve the reproducibility of the region separated into character regions, particularly black characters or color characters. Binarization processing or multilevel processing is performed.

  In addition, the gradation reproduction processing unit 29 performs gradation reproduction processing so that the region separated by the halftone dot region in the region separation processing unit 23 can be finally separated into pixels and the respective gradations can be reproduced. (Halftone generation) is performed. Further, the gradation reproduction processing unit 29 binarizes or multi-values the region separated into the photographic region by the region separation processing unit 23 on the screen with an emphasis on gradation reproducibility in the color image output device 3. To do.

  The filter coefficient setting unit 30 sets a filter coefficient set used in the spatial filter processing unit 26. The setting in the filter coefficient setting unit 30 is a main feature of the present invention, and details thereof will be described later.

  The color image processing apparatus 2 temporarily stores the image data (CMYK signal) processed by the gradation reproduction processing unit 29 in a storage unit (not shown), and performs a predetermined timing for image formation of the image data stored in the storage unit. The read image data is output to the color image output device 3. These controls are performed by, for example, a CPU or a DSP (both not shown).

  The operation panel 4 includes, for example, a display unit such as a liquid crystal display and setting buttons. The color image input device 1, the color image processing device 2, and the color image output device 3 are based on information input from the operation panel 4. Operation is controlled.

  In the image forming apparatus 100 according to the present embodiment, the scaling unit 27 of the color image processing apparatus 2 changes the image data acquired by reading the document by the color image input apparatus 1 or the image data received from the outside. In addition to performing the multiplication process, the spatial filter processing unit 26 performs the spatial filter process based on the filter coefficient set corresponding to the scaling process. As a result, even when scaling processing is performed on the image data at different scaling ratios in the main scanning direction and the sub-scanning direction, a high-quality original can be output. Note that the image data to be subjected to the scaling process is not limited to the data read from the original by the color image input device 1 such as received from the outside as described above.

  In the description regarding the spatial filter processing unit 26 and the scaling unit 27 described above, it has been described on the assumption that the processing in the scaling unit 27 is after the processing in the spatial filter processing unit 26. The process may be configured to be executed before the process in the spatial filter processing unit 26. In FIG. 1, the filter processing is performed only by the spatial filter processing unit 26, but another filter processing unit can be added at an appropriate position before or after the filter processing.

  Next, as main features of the present invention, details of the filter coefficient set used in the spatial filter processing unit 26, a detailed configuration of the filter coefficient setting unit 30, and details of processing performed by the filter coefficient setting unit 30 will be described. . In the following description, symbols C, M, and K are used. Of course, these symbols are different from symbols representing colors in the CMYK signal.

  First, a general configuration of a filter used in the spatial filter processing unit 26 will be described with reference to FIG. The filter 50 shown in FIG. 2 is a filter coefficient set whose vertical and horizontal sizes are (2Ms + 1) × (2Ss + 1), and is stored in the color image processing apparatus 2 in advance. Note that the filter 50 and the filters whose sizes are defined by M and S, which will be described later, and M0 and S0, may be the same size or different sizes in the vertical and horizontal directions. The vertical and horizontal sizes are preferably odd numbers. Such an example is given in the filter 50 of FIG.

  In the filter 50, the numerical values that can be taken by the filter coefficients a [0, 0] to a [Ms, Ss] are rational numbers having a common denominator or rational numbers having a numerator and a denominator set independently. In the example of FIG. 2, the filter coefficient a [0, 0] located at the center corresponds to the filter coefficient for the pixel of interest, and the filter coefficients a [0, 0] to a [Ms, 0] and the filter coefficient a [0, 0]. ˜a [0, Ss] corresponds to the axes (vertical axis and horizontal axis, respectively). Then, by setting the filter coefficients symmetrically with respect to these axes, it is not necessary to set and store the hatched portion coefficients in FIG. 2 separately from the white background portion and halftone portion coefficients. .

Next, a configuration example of the filter coefficient setting unit 30 will be described with reference to FIG.
The filter coefficient setting unit 30 illustrated in FIG. 3 includes a first conversion matrix setting unit 301, a second conversion matrix setting unit 302, a filter coefficient set selection unit 303, and coefficient conversion units 304a and 304b. In FIG. 3 and FIG. 4 to be described later, the parts with a and b are distinguished as performing the processes corresponding to the first transformation matrix setting unit 301 and the second transformation matrix setting unit 302, respectively. When both are described collectively, the description may be made by omitting a and b from the reference numerals.

  The first conversion matrix setting unit 301 sets the first conversion matrix according to the vertical scaling factor when the scaling process is performed on the image data. The second transformation matrix setting unit 302 sets the second transformation matrix according to the scaling factor in the horizontal direction (direction perpendicular to the vertical direction) when the scaling process is performed on the image data. The first and second conversion matrices are preferably set by calculation, but may be set by reading out from previously stored conversion matrices according to the scaling factors in the corresponding directions. The zoom ratio in the vertical direction and the zoom ratio in the horizontal direction are zoom ratios designated for the main scanning direction and the sub-scanning direction, respectively, and are hereinafter referred to as a main scanning direction scaling ratio and a sub-scanning direction scaling ratio.

  The filter coefficient set selection unit 303 is a selection unit that selects a reference filter coefficient set based on (i) image processing conditions and (ii) one or more filter coefficient sets stored in advance. Here, the image processing condition is either (i-1) a predetermined image processing condition different from the scaling factor in the main scanning direction and the scaling factor in the sub-scanning direction, or (i-2) the main processing condition under the predetermined image processing condition. This refers to image processing conditions including a magnification in the scanning direction and a magnification in the sub-scanning direction. Examples of the predetermined image processing conditions include the degree of smoothing, the degree of edge enhancement, and the document type. The image processing conditions are not limited to those designated from the operation panel 4 by the user, and may be determined based on the result of automatically reading the document. Of course, a certain default reference filter coefficient set may be selected under any image processing conditions.

  The coefficient conversion unit 304 a is set by the first conversion matrix setting unit 301 as a matrix (matrix generated by arranging the reference filter coefficient sets) indicated by the reference filter coefficient set selected by the filter coefficient set selection unit 303. A linear transformation is performed using the first transformation matrix. The coefficient conversion unit 304b performs primary conversion on the matrix indicated by the reference filter coefficient set selected by the filter coefficient set selection unit 303 using the second conversion matrix set by the second conversion matrix setting unit 302. Actually, the coefficient conversion unit 304a performs primary conversion on the filter coefficient set after the primary conversion using the second conversion matrix and outputs the result to the spatial filter processing unit 26, or vice versa. The conversion may be performed according to the procedure and output from the coefficient conversion unit 304a to the spatial filter processing unit 26. Hereinafter, the former case will be described.

  Note that the filter coefficient setting unit 30 may have a configuration in which one of the coefficient conversion unit 304a and the coefficient conversion unit 304b is omitted. Further, in such a configuration, the conversion matrix setting unit corresponding to the omitted side (main scanning direction or sub-scanning direction) is also omitted, and a reference filter coefficient set that matches such omission is selected. It can also be configured.

  As described above, in the color image processing apparatus 2, the coefficient conversion unit sets the matrix indicated by the reference filter coefficient set selected by the filter coefficient set selection unit 303 as the first conversion matrix setting unit 301 and / or the second conversion matrix setting. The linear transformation is performed using the first transformation matrix and / or the second transformation matrix set by the unit 302. Referring to the example of FIG. 2, after creating a (2Ms + 1) × (2Ss + 1) reference filter coefficient set using symmetry based on the selected (Ms + 1) × (Ss + 1) reference filter coefficient set, Is linearly transformed, or the selected (Ms + 1) × (Ss + 1) reference filter coefficient set is linearly transformed, and a (2Ms + 1) × (2Ss + 1) filter coefficient set is created using symmetry.

  Then, the spatial filter processing unit 26 performs filter processing on the image data using the filter coefficient set after the primary conversion by the coefficient conversion unit 304a and / or the coefficient conversion unit 304b. However, in the present invention, since the spatial filter processing based on the filter coefficient set corresponding to the scaling process is performed by the spatial filter processing unit 26, the coefficient conversion unit performs the first conversion when the vertical and horizontal scaling factors are different. The linear transformation is performed using both the matrix and the second transformation matrix.

  The color image processing apparatus 2 according to the present invention performs primary conversion of the reference filter coefficient set using the two conversion matrices corresponding to the two directions as described above, and performs filter processing using the filter coefficient set obtained thereby. As a result, an optimized filter coefficient set is generated in consideration of the frequency characteristics of both the main scanning direction and the sub-scanning direction of the image data even when the scaling ratios in the main scanning direction and the sub-scanning direction are different. The image processing time can be shortened because the image processing can be performed by a single filtering process.

Next, another configuration example of the filter coefficient setting unit 30 will be described with reference to FIG.
In the configuration example described here, in the configuration example of FIG. 3, the number of rows and columns that can be processed by the spatial filter processing unit 26 is determined (or the number of rows and columns to be processed is limited). ). The details of the configuration example of FIG. 4 will be described below, but the details of the parts included in the configuration example of FIG. 3 can also be basically used in this description.

  The actual size to be converted is indicated by (2M + 1) × (2S + 1) in FIG. Also, the coefficients used for the number of rows x the number of columns that can be processed by the spatial filter processing unit 26 are indicated by halftone dots in FIG. That is, in the spatial filter processing unit 26 in this configuration example, the filter processing is finally performed with a filter coefficient set having a size of M0 × S0 in which the halftone dot portions (locations indicated by the halftone dots) are combined. . In FIG. 2, it is assumed that M0 ≦ Ms, M ≦ Ms, S0 ≦ Ss, and S ≦ Ss for the above sizes.

  The filter coefficient setting unit 30 illustrated in FIG. 4 includes size correction units 305a and 305b, filter coefficient combination units 306a and 306b, a window function processing unit 307, and a division coefficient calculation unit 308 in addition to the configuration example of FIG. ing. In addition, the filter coefficient setting unit 30 illustrated in FIG. 4 can omit any one of the parts 301 and 304a to 306a and the parts 302 and 304b to 306b.

  In FIG. 4, the size correction units 305a and 305b, the filter coefficient combination units 306a and 306b, the window function processing unit 307, and the division coefficient calculation unit 308 can be configured without any of them. The order of the size correction units 305a and 305b, the filter coefficient combination units 306a and 306b, the window function processing unit 307, and the division coefficient calculation unit 308 can be other than that shown in FIG. 4, but the order shown in FIG. This is the desired order.

  First, the scaling factor in the scaling process to be performed by the scaling unit 27 of the color image processing apparatus 2, such as the scaling factor specified by the user via the operation panel 4, is set in the first conversion matrix of the filter coefficient setting unit 30. Input to the unit 301 and the second transformation matrix setting unit 302. Note that the image forming apparatus 100 according to the present embodiment is configured so that different magnifications can be specified (set) for each of the main scanning direction and the sub-scanning direction. One of the magnifications is input to the first transformation matrix setting unit 301, and the other one is input to the second transformation matrix setting unit 302. Actually, it depends on the reading resolution (first resolution) in the main scanning direction and the reading resolution (second resolution) in the sub-scanning direction specified by the user or the default scaling factor in the main scanning direction and sub-scanning direction specified by the user. The actual main scanning direction scaling factor and sub-scanning direction scaling factor are determined. In the following description, it is assumed that the first transformation matrix setting unit 301 sets the main scanning direction scaling factor and the second transformation matrix setting unit 302 sets the sub scanning direction scaling factor.

  The first conversion matrix setting unit 301 and the second conversion matrix setting unit 302 respectively input a matrix M (x ′, x) corresponding to the input main scanning direction scaling factor and a matrix S (y) corresponding to the sub scanning direction scaling factor. ′, Y) is calculated and output to the coefficient conversion unit 304a and the coefficient conversion unit 304b. This calculation method will be described later.

  In the memory (storage unit) 300, the filter 50 of FIG. 2, that is, (2Ms + 1) × (2Ss + 1) filter coefficients is stored in advance for extraction of a reference filter coefficient set. The memory 300 may be provided in the filter coefficient setting unit 30, may be provided in the color image processing device 2 outside the filter coefficient setting unit 30, or may be provided in the color image processing device 2. It may be provided in the external image forming apparatus 100.

  The filter coefficient set selection unit 303 selects a reference filter coefficient set A (x, X) from the filter coefficient set stored in the memory 300 based on the main scanning direction scaling ratio, the sub scanning direction scaling ratio, and other predetermined image processing conditions. y) Select and read. As described above, since the filter coefficient is symmetric with respect to the main axis (vertical axis) and the sub-axis (horizontal axis) passing through the pixel of interest, the coefficient in the hatched portion may use the coefficient of the filter coefficient set and is stored in the memory 300. There is no need to do. The filter coefficient set selection unit 303 outputs the reference filter coefficient set A (x, y) read from the memory 300 to the coefficient conversion unit 304a.

The coefficient conversion unit 304a performs primary conversion on the reference filter coefficient set A (x, y) using the matrix M (x ′, x) input from the first conversion matrix setting unit 301 and outputs the result to the coefficient conversion unit 304b. . The coefficient conversion unit 304b performs primary conversion using the matrix S (y ′, y) input from the second conversion matrix setting unit 302, and the obtained filter coefficient set C (x ′, y ′) is spatially converted. Output to the filter processing unit 26. The above calculations are collectively expressed as in equation (1).
C = M * A * S ^T (1)

  By the way, in the primary conversion in the coefficient conversion unit 304, the size M + 1 of the matrix M (x ′, x) and the size S + 1 of the matrix S (y ′, y) are the size (Ms + 1) × (Ss + 1) of A (x, y). And may be different. At this time, the rows of A (x, y) converted by the matrix M (x ′, x) are 0 to M, and the columns of A (x, y) converted by the matrix S (y ′, y) are 0. ˜S, and M + 1 rows and after and S + 1 columns and after are not subject to conversion. However, this processing is equivalent to the case where the size of the matrix M (x ′, x) and the matrix S (y ′, y) is set to (Ms + 1) × (Ss + 1) with the setting of not converting M + 1 rows and later and S + 1 columns and later. Yes, it can also be expressed by the formula (1) on the calculation formula. Accordingly, with regard to the primary conversion, the subsequent description will proceed with the sizes of the matrices M (x ′, x) and S (y ′, y) as Ms + 1 and Ss + 1, respectively.

  The matrices M (x ′, x) and S (y ′, y) are square matrices that are calculated according to the setting of image processing conditions such as a scaling factor. The reference filter coefficient set A (x, y) is also a matrix selected in accordance with the setting of image processing conditions such as a scaling factor. Any of the image processing conditions can include the predetermined image processing conditions. As described above, the matrices M (x ′, x) and S (y ′, y) have at least the main scanning direction scaling factor and the sub-scanning. The reference filter coefficient set A (x, y) is selected according to at least other predetermined image processing conditions.

  Furthermore, the scaling factor used for selecting the reference filter coefficient set A (x, y), the matrix M (x ′, x), and S (y ′, y) is a variable that can be switched continuously. Therefore, the reference filter coefficient set A (x, y) and the matrices M (x ′, x) and S (y ′, y) corresponding to the main scanning direction scaling factor and the sub-scanning direction scaling factor have these scaling factors. Each is divided into a plurality of sections, and selected according to each section.

  The filter coefficient setting unit 30 is not limited to the example using only the scaling factor, and the filter coefficient setting unit 30 changes the scaling factor in the main scanning direction and the sub-scanning direction according to the conditions included in the predetermined image processing condition or the predetermined image processing condition. The conditions included in the image processing conditions to which the magnification is added are preferably divided into a plurality of sections, and the first conversion matrix, the second conversion matrix, and / or the reference filter coefficient set are switched in each section. Thereby, the filter at the time of filter processing can be finely switched for every condition included in the image processing condition.

  At this time, the reference filter coefficient set A (x, y), the matrix M (x ′, x), and S (y ′, y) in two adjacent sections are divided into three variables on the boundary of the section and between them. It is preferable that the difference in effect on the image in the filter processing of the filter coefficient set C (x ′, y ′) at the magnification is set to be small. That is, A (x, y), matrix M (x ′, x), and S (y ′, y) in each section are reduced so that the error of the spatial frequency characteristic of the filter coefficient generated in each section is reduced. It is preferable to set. Note that the reference filter coefficient set A (x, y), matrix M (x ′, x), and S (y ′, y) on the boundary are either one of the sections on both sides or the sections on both sides. It is calculated from any one of these.

  Here, since each condition is divided into the plurality of sections, for example, when there are two conditions P and Q whose values continuously change, the condition Q is divided into Np sections for the condition P. Is divided into a total of Np × Nq, and the reference filter coefficient set A (x, y) and / or the matrix M (x ′, x), S (y ′, y) may be set. However, the condition P and the condition Q usually differ in how to divide the divisions and set values on the boundaries.

  In this configuration example, the number of rows Ms and the number of columns Ss of the reference filter coefficient set A (x, y) may be larger than the number of rows M0 and the number of columns N0 possible in the spatial filter processing unit 26. However, in this case, the number of rows and the number of columns of the converted filter coefficient set C (x ′, y ′) are the same as those of the reference filter coefficient set A (x, y). It cannot be used for processing.

  Therefore, a size correction unit 305 is provided, and the size correction unit 305 corrects the number of rows to M0 or less and the number of columns to S0 or less. An example of the filter size reduction processing in the size correction unit 305a or 305b at this time is shown in FIG. The filter coefficients are deleted in order from the target pixel to the Mth row and the Sth column, starting with the row and column closer to the target pixel.

  The processing in the row direction at that time is performed by the size correcting unit 305a as follows. First, after outputting Ms ≧ m> M rows to the filter combination unit 306a (step S10), the remaining M + 1 rows are unconditionally deleted until the number of rows reaches M0 + M−Ms (step S11). If YES, step S12). When the number of rows becomes M0 + M−Ms (NO in step S11), the sum of the filter coefficients in the row farthest from the target pixel is calculated (step S13). All are replaced with 0 (in the case of YES in step S14, step S15). This process is repeated until the sum of the filter coefficients in the row farthest from the target pixel does not exceed the threshold value (NO in step S14). Note that the threshold value at this time is read because it is stored in the memory 300, and the value may be constant for each line or may be set for each line. The processing in the column direction is similarly performed in the size correcting unit 305b that is different from the row direction. In FIG. 5, “Mth row” is changed to “Sth column”, “Ms” is changed to “Ss”, “M” is changed to “S”, “M0” is changed to “S0”, and “306a”. Is replaced with “306b”, and the description is omitted.

  In this way, the size of the filter coefficient set obtained by linearly transforming the reference filter coefficient set A (x, y) is corrected by the size correction units 305a and 305b. In this way, the size correction units 305a and 305b convert the filter coefficient set after the primary conversion into the size of the reference filter coefficient set or a predetermined size (of course, the predetermined size of the reference filter set). Adjust the size so that it fits. Therefore, the size correction units 305a and 305b can be said to be size adjustment units that perform such adjustment. In the processing procedure of FIG. 4, the size correcting unit 305a outputs the corrected filter coefficient set to the filter coefficient combining unit 306a and combines it with the coefficients from the (M + 1) -th row output previously. The size correcting unit 305b outputs the corrected filter coefficient set to the filter coefficient combining unit 306b and combines it with the coefficients of the (S + 1) th and subsequent lines output previously.

  Further, the window function processing unit 307 can input the output from the filter coefficient combining unit 306b and multiply the input filter coefficient set by a matrix having the same size as the filter calculated by the window function for each element. It is. FIG. 6 shows a matrix (filter coefficient set) 60 as an example of a filter used for window function processing in such a window function processing unit 307. The matrix 60 becomes a Hanning window in the main scanning direction and the sub-scanning direction when a filter composed of, for example, 5 × 5 filter coefficient sets is input from the filter coefficient combining unit 306b, and the filter is multiplied. This is an example of a simple matrix.

  Finally, the division coefficient calculation unit 308 receives the output from the window function processing unit 307 and calculates a division coefficient for the input filter coefficient set. Here, the calculation is generally performed by calculating the sum of the coefficients to obtain the division coefficient so that the density preservation is maintained even if the filter coefficient is changed. However, the present invention is not limited to this.

  The series of operations as described above are executed when the image processing conditions are changed, such as changing the magnification. However, when calculating the conversion matrix corresponding to the initial scaling ratio that is initially set when the color image processing apparatus 2 is started, an operation that requires filter processing first is started from the start of the color image processing apparatus 2. It can also be executed during any of the above.

  The size correction unit 305, the filter coefficient combination unit 306, the window function processing unit 307, the division coefficient calculation unit 308, or any one or both of the parts 304a to 306a and parts 304b to 306b are operated by the user operation. It is preferable that the configuration can be omitted. Further, when there is no change in the setting related to the calculation of the conversion matrix such as the scaling factor and the reading resolution since the previous processing, the conversion by the first conversion matrix setting unit and / or the second conversion matrix setting unit is omitted. It is preferable to be configured as follows. Further, any one of the filter coefficient set obtained from the process in the coefficient conversion unit 304 to the process in the division coefficient calculation unit 308 and the filter coefficient set selected in the filter coefficient set selection unit 303 is operated by the user. Thus, it can be selected as a filter coefficient set used for actual filter processing.

  Next, details of calculation of the matrices M (x ′, x) and S (y ′, y) in the first transformation matrix setting unit 301 and the second transformation matrix setting unit 302 will be described. In the matrix display, the vertical scanning is the main scanning and the horizontal scanning is the sub-scanning.

The characteristics of the filter used for the filter processing are generally expressed by Fourier transform. This can be expressed in the form of primary transformation as shown in equation (2-1).
X1 = FM * F * FS ^T (2-1)

  Here, F is a filter, FM (u, x) is a matrix of 2Ms + 1 rows and 2Ms + 1 columns with exp (-2πi * u * x / (2 * Ms + 1)) as elements of u rows and x columns, and FS (v, y). ) Is a 2Ss + 1 row 2Ss + 1 column matrix with exp (−2πi * v * y / (2 * Ss + 1)) as an element of v rows and y columns, and X1 (u, v) is a matrix of 2Ms + 1 rows Ss + 1 columns. At this time, integer values that u and x can take are -Ms to + Ms, and integer values that v and y can take are -Ss to + Ss. Note that T in the upper right of the matrix indicates transposition, that is, replacement of rows and columns.

Furthermore, as shown in FIG. 2, the filter coefficients used for image processing are generally symmetric with respect to the main axis and the sub-axis passing through the target pixel. In this case, the Fourier transform X1 (u, v) of the reference filter coefficient set A (x, y) generated using the filter elements can be expressed only by the cos function as shown in Expression (2-2). it can.
X1 = CM * A * CS ^T (2-2)

Here, CM (u, x) is a matrix of Ms + 1 rows and Ms + 1 columns with cos (-2π * u * x / (2 * Ms + 1)) as elements of u rows and x columns, and CS (v, y) is cos This is a matrix of Ss + 1 rows and Ss + 1 columns with (−2π * v * y / (2 * Ss + 1)) as elements of v rows and y columns. At this time, integer values that u and x can take are 0 to Ms, and integer values that v and y can take are 0 to Ss. A (x, y) is calculated by the equation (2-3).
A = F (0: Ms, 0: Ss) (2-3)

  Here, F (0: Ms, 0: Ss) is obtained by extracting u = 0 to Ms rows and v = 0 to Ss columns of the filter F.

By the way, cos (−2π * u * x / (2 * Ms + 1)) and cos (−2π * v * y / (2 * Ss + 1)) are respectively cos (−2π * u / (2 * Ms + 1)), Cos (−2π * v / (2 * Ss + 1)) Chebyshev polynomial. Therefore, CM (u, x) and CS (v, y) can be converted as shown in Equation (3).
CM = DM * T, CS = DS * T (3)

  Here, T (x1, x) and T (y1, y) are triangular matrices composed of only integers, and the elements are as in equation (4-1). In the formula (4-1), m represents x1 or y1, and n represents x or y.

When m, n ≦ M or S, T (m, n) = K (m, n) * T0 (m, n)
T0 (0,0) = 1, T0 (0,1) = 0, T0 (1,1) = 1
T0 (m, n) = 2 * T0 (m-1, n-1) -T0 (m, n-2) (m≤n)
T0 (m, n) = 0 (m> n)
In other cases, T0 (m, n) = 1 (m = n), T0 (m, n) = 0 (m ≠ n)
... (4-1)

K (m, n) is an integer value. There is no particular restriction as to what integer is multiplied, but generally the following equation (4-2) is used.
K (m, n) = 1 (n = 0), 2 (others) (4-2)

Further, DM (u, x1) in the x1 ≦ M [cos (-2π * u / (2 * Ms + 1))] and ^x1 is an element of u rows x1 column, x1> in M CM (u, x1) Is a matrix of Ms + 1 rows and Ms + 1 columns with elements of u rows and x1 columns, DS (v, y1) is [cos (−2π * v / (2 * Ss + 1))] for ^y1 ≦ S, and ^y1 is of v rows and y1 columns It is a matrix of Ss + 1 rows and Ss + 1 columns with elements of and CS (v, y1) as elements of v rows and y1 columns when y1> S.

By using Expression (3), Expression (2-2) can be expressed in the form of Expression (5).
X1 = DM * T * A * T ^T * DS ^T (5)

Furthermore, Expression (3) is expressed as Expression (6). Here, CM is [1-cos (-2π * u / (2 * Ms + 1))] when x2 ≦ M, and ^x2 is an element of u rows and x2 columns, and CM (u, x2) is u when x2> M. Expressed using a matrix EM (u, x2) of Ms + 1 rows Ms + 1 columns as elements of rows x2 columns, CS is [1-cos (−2π * v / (2 * Ss + 1))] ^{y2 when} y2 ≦ S. Is represented by a matrix ES (v, y2) of Ss + 1 row Ss + 1 column, where CS (v, y2) is an element of v row y2 column, where y is an element of v row y2 column.
CM = DM * T = EM * P * T, CS = DS * T = ES * P * T (6)

  Here, P (y2, y1) is a triangular matrix composed of only integers, and its elements are as shown in Equation (7). In the formula (7), m represents x2 or y2, and n represents x1 or y1.

When m, n ≦ M or S, P (0,1) = 1
P (m, n) = − P (m−1, n−1) + P (m, n−1) (m ≦ n)
P (m, n) = 0 (m> n)
Otherwise, P (m, n) = 1 (m = n), P (m, n) = 0 (m ≠ n)
... (7)

By using Expression (6), Expression (5) can be expressed in the form of Expression (8-1).
X1 = EM * P * T * A * T ^T * P ^T * ES ^T (8-1)

Next, a method for enlarging / reducing the frequency characteristic in the main scanning direction and the sub-scanning direction will be described based on the equations (2-1) to (8-1).
Expression (8-1) is expressed in the form of a polynomial that is 0 when the element at x2 ≦ M of EM is u = 0, and 0 when the element at y2 ≦ S of ES is v = 0. Yes. Therefore, by multiplying [1-cos (-2π * u / (2 * Ms + 1))] and [1-cos (−2π * v / (2 * Ss + 1))] by appropriate numerical values γM and γS, respectively. The frequency characteristics can be enlarged or reduced in the main scanning direction and the sub-scanning direction. The calculation at this time can be expressed as in Expression (9).
X2 = EM * RM * P * T * A * T ^T * P ^T * RS * ES ^T (9)

Here, RM and RS are diagonal matrices composed of a geometric sequence with a diagonal component of γ, and the diagonal components are expressed as in Equation (10).
When m ≦ M or S, RM (x2, x2) = γM ^u−1 , RS (y2, y2) = γS ^v−1
Otherwise, RM (x2, x2) = 1, RS (y2, y2) = 1
(10)

  γM and γS can be set in the form of an LUT (Look Up Table) or the like according to the resolution and the scaling factor. These elements are usually set to increase as the reading resolution increases or as the user-specified scaling factor decreases.

On the other hand, the Fourier transform of the filter coefficient set C (x ′, y ′) after enlargement / reduction can be expressed in the same manner as Expression (8-1).
X2 = EM * P * T * C * T ^T * P ^T * ES ^T (8-2)

From the expressions (8-2) and (9), the conversion of the filter coefficient can be expressed as the expression (11).
C = (P * T) ⁻¹ * RM * P * T * A * T ^T * P ^T * RS * (T ^T * P ^T ) ⁻¹
(11)

When Expression (1) and Expression (11) are compared, Expression (12) is obtained.
M = (P * T) ⁻¹ * RM * P * T, S = (P * T) ⁻¹ * RS * P * T
(12)

  At this time, P and T are triangular matrices, and RM and RS are diagonal matrices as shown in Equation (10), so M and S, which are products of diagonal matrices and triangular matrices, are triangular matrices.

Further, the equation (12) can be expressed by being transformed as the equation (13).
RM = (P * T) * M * (P * T) ⁻¹ , RS = (P * T) * S * (P * T) ⁻¹
... (13)

  Since RM and RS are diagonal matrices, equation (13) also indicates that the first and second transformation matrices are diagonalized by a triangular matrix consisting only of integers. At that time, the diagonal matrix is represented by a geometric sequence. In other words, the first and second transformation matrices M and S exemplified here are matrices that are diagonalized by a triangular matrix consisting only of integers, and the diagonal components of the matrix obtained by diagonalization are rational numbers. It can be approximated by a geometric sequence.

  Now, the equation (4-1) in the above description is a recurrence equation used for generating the Chebyshev polynomial. The Chebyshev polynomial is a kind of polynomial called an orthogonal polynomial, and is an expression often used in filter design.

  However, in the present invention, it is possible to use not only the Chebyshev polynomial but also other polynomials, particularly orthogonal polynomials. For example, when a Legendre polynomial is used instead of the Chebyshev polynomial, the equation (4-1) becomes the equation (4-3). Here, in any of the equations (4-1) and (4-3), it is not necessary to limit the number of elements other than 1 row and 1 column of K to 2.

When m, n ≦ M or S, T (m, n) = K (m, n) * T0 (m, n)
T0 (0,0) = 1, T0 (0,1) = 0, T0 (1,1) = 1
T0 (m, n + 1) = (2-1 / n) * T0 (m-1, n)-(1-1 / n) * T0 (m, n-2) (m≤n)
T0 (m, n) = 0 (m> n)
In other cases, T0 (m, n) = 1 (m = n), T0 (m, n) = 0 (m ≠ n)
... (4-3)

  Even if the other polynomial is used instead of the equation (4-1), since the T is a triangular matrix, the equations (5) to (12) can be applied as they are. M and S corresponding to the first and second transformation matrices are also triangular matrices.

  Thus, the first transformation matrix is a triangular matrix consisting of only rational numbers generated by a predetermined calculation using one variable based on the scaling factor in the main scanning direction, and the second transformation matrix is the sub-scanning direction. It is preferable that it is a triangular matrix consisting only of rational numbers generated by a predetermined calculation using one variable based on the scaling factor. As a result, the spatial frequency characteristic of the filter used for the filter processing becomes suitable for a predetermined scaling factor. Further, as described above, the coefficient of the orthogonal polynomial or the coefficient of the Chebyshev polynomial can be used for the calculation of the triangular matrix.

  Next, an example of the operation of the color image processing apparatus 2 will be described with reference to filter conversion examples. 7 to 10 are diagrams for explaining an example of the filter coefficient setting process in the filter coefficient setting unit in the image forming apparatus of FIG. 7 is a diagram illustrating an example of a reference filter coefficient set, FIGS. 8A to 8C are diagrams illustrating examples of transformation matrices, and FIGS. 9A to 9C are diagrams illustrating the reference filter coefficient sets of FIG. FIG. 10 is a diagram showing frequency coefficient in the main scanning direction of the filter coefficient set of FIGS. 9A to 9C.

  11 and 12 are flowcharts for explaining an example of the filter coefficient setting process in the filter coefficient setting unit 30, FIG. 11 shows an example of the conversion process in the main scanning direction of the filter, and FIG. 11 shows an example of conversion processing in the sub-scanning direction of the filter executed following the processing of No. 11. The filter coefficient setting process is not only configured by a dedicated hardware circuit such as the filter coefficient setting unit 30 but also a filter coefficient setting process procedure is determined for a personal computer including a CPU, a RAM, a ROM, and the like. It can also be performed by loading a computer program and causing it to be executed by a CPU (both not shown).

  When a document is copied by the image forming apparatus 100, image data acquired by the color image input apparatus 1 reading the document is input to the color image processing apparatus 2, and the main scanning direction set by the user via the operation panel 4 is used. The magnification ratio and the sub-scanning direction scaling ratio, or the main scanning direction scaling ratio and the sub-scanning direction scaling ratio based on the scanning speed of the color image input apparatus 1 are input to the color image processing apparatus 2. The image data may be acquired by reading a document with the color image input device 1 or may be acquired by receiving electronic data created by a processing device such as a personal computer.

  11 and 12, the filter coefficient setting unit 30 first sets a reference filter coefficient set according to the main scanning direction scaling factor and the sub scanning direction scaling factor from one or a plurality of filter coefficient sets set in the memory 300. Is selected (step S21). Here, for example, the reference filter coefficient set 70 of FIG. 7 is selected.

  Next, it is determined whether or not the inputted main scanning direction magnification is equal (Step S22). When it is determined that the main scanning direction magnification is not equal (NO in step S22), the filter coefficient setting unit 30 obtains a first conversion matrix (main scanning direction conversion matrix) corresponding to the main scanning direction magnification. The calculation is performed (step S23), and the reference filter coefficient set 70 is converted by the main scanning direction conversion matrix (step S24). In step S23, for example, any one of transformation matrices 81 to 83 shown in FIGS. 8A to 8C is calculated. Note that all of the transformation matrices 81 to 83 are rational numbers. Further, although the transformation matrices 81 to 83 are shown using symmetry as in the filter 50 of FIG. 2, in step S24, the transformation matrix is converted into 4 × 4 elements (coefficients) in the lower right of the reference filter coefficient set 70. Any one of 81 to 83 may be multiplied and copied so as to be symmetrical.

  Whether or not the converted filter coefficient set is used in the spatial filter processing unit 26 is selected based on a user operation. This user operation can be set and registered in advance. If use is selected (YES in step S25), the next step is executed on the converted filter coefficient set. When the use is not selected (NO in step S25) and when the main scanning direction magnification is determined to be equal (when YES in step S22), the following is performed for the reference filter coefficient set 70. A step is executed. Note that step S25 may be omitted.

  Next, it is determined whether or not the input sub-scanning direction variable magnification is equal (step S31). If it is determined that the sub-scanning direction magnification is not equal (NO in step S31), the filter coefficient setting unit 30 obtains a second conversion matrix (sub-scanning direction conversion matrix) corresponding to the sub-scanning direction magnification. The filter coefficient set or reference filter coefficient set 70 converted in step S24 is converted by the sub-scanning direction conversion matrix (step S33). In step S32, as in step S23, for example, any one of transformation matrices 81 to 83 shown in FIGS. 8A to 8C is calculated. Here, for simplification of description, an example in which the same conversion matrix is used in the main scanning direction and the sub-scanning direction will be described, but the present invention is not limited thereto.

  When the same conversion matrix is used among the conversion matrices 81 to 83 in the main scanning direction and the sub scanning direction, the filter coefficient sets converted in step S33 are the filter coefficient sets 91 to 91 shown in FIGS. 9A to 9C, respectively. 93.

  Whether or not the converted filter coefficient set is used in the spatial filter processing unit 26 is selected based on a user operation. This user operation can be set and registered in advance. If use is selected (YES in step S34), the converted filter coefficient set is output to the spatial filter processing unit 26, and the process ends. When the use is not selected (NO in step S34) and when the sub-scanning direction variable magnification is determined to be equal (if YES in step S31), the filter coefficient output in the process of FIG. The set is output to the spatial filter processing unit 26 as it is, and the process ends.

  Here, the division coefficient of the reference filter coefficient set 70 shown in FIG. 7 is 49, and the division coefficients of the converted filters 91, 92, and 93 shown in FIGS. 9A, 9B, and 9C are 49, 45, and 49, respectively. Become. In practice, the division coefficient calculation unit 308 corrects the filter coefficient set based on these division coefficients, and then outputs them to the spatial filter processing unit 26.

  FIG. 10 shows the frequency characteristics of the filter coefficient sets 91 to 93 in the main scanning direction. If the first conversion matrix and the second conversion matrix are made the same, basically the same frequency characteristics can be obtained in the sub-scanning direction. The conversion matrices 82 and 83 are triangular matrices, and the frequency characteristics corresponding to them are indicated by a broken line B and a broken line C, respectively. A frequency characteristic corresponding to the conversion matrix 81 which is a unit matrix is indicated by a solid line A.

  When these are compared, the use of a triangular matrix rather than a unit matrix as the first and second transformation matrices makes the spatial frequency characteristics of the filter used for the filter processing more suitable for larger scaling factors or lower resolutions. I understand that. In the case of FIGS. 9A to 9C, referring to FIG. 10, since the broken line C has a higher intensity at a high frequency, the output image in the color image output device 3 is less blurred and suitable for a case where the magnification is large. Yes. On the other hand, since the solid line A has low intensity at high frequencies, moire caused by high frequency components is less likely to occur when the output image is reduced in the color image output device 3, and is suitable for a case where the magnification is small. Yes.

  As described above, in the present embodiment, both the main scanning direction and the sub-scanning direction of the image data are obtained even when the scaling factor of the scaling process performed on the image data is different between the main scanning direction and the sub-scanning direction. It is possible to generate one set of filter coefficients that is optimized in consideration of the frequency characteristics (that is, in consideration of the respective scaling factors). Therefore, the spatial filter processing unit 26 can perform a filter process that realizes high image quality of the output image based on one filter coefficient set optimized by the filter coefficient setting unit 30. The number of filter coefficient sets can be reduced, and the amount of calculation and the required memory capacity can be reduced.

  Further, in the present invention, in actuality, the filter coefficient set is calculated without performing Fourier transform or inverse Fourier transform, and the spatial filter processing based on the filter coefficient set is performed. Therefore, when returning from the spatial frequency component to the filter coefficient set (reverse after Fourier transform). There is no problem that non-optimal filter coefficients are generated due to errors or bugs that occur during Fourier transformation. Furthermore, by using a matrix consisting of only integers or rational numbers for all elements in the reference filter coefficient set and the first and second transformation matrices, the optimum filter coefficients can be generated more reliably by calculation.

  The image forming apparatus 100 described with reference to FIG. 1 is not limited to a digital color copying machine, but may be applied to a digital color multifunction peripheral having a copier function, a printer function, a facsimile transmission / reception function, a scan to e-mail function, and the like. . The digital color multi-function peripheral can further include a communication device such as a modem or a network card. In this case, for example, when transmitting a facsimile, if the modem performs a transmission procedure with the other party and secures the transmission possible state, the image data compressed in a predetermined format (the image read by the scanner) Data) is read from the memory, subjected to necessary processing such as changing the compression format, and sequentially transmitted to the other party via a communication line.

  When receiving a facsimile, the image data transmitted from the other party is received while performing a communication procedure, and is output to the color image processing apparatus 2. The color image processing apparatus 2 receives the received image data (not shown). The compression / decompression processing unit performs decompression processing. The decompressed image data is subjected to rotation processing and resolution conversion processing as necessary, subjected to output tone correction and tone reproduction processing, and is output from the color image output device 3.

  Further, the image forming apparatus 100 may be configured to perform data communication with a computer and other digital multi-function peripherals connected to the network via a network card, a LAN cable, or the like. Further, the present invention can be applied not only to a color multifunction peripheral but also to a multifunction peripheral that handles monochrome images and a facsimile communication apparatus that has only a facsimile communication function.

  In the above-described embodiment, as the color image input device 1, for example, a flat bed scanner, a film scanner, a digital camera, a mobile phone, or the like is used. Instead of acquiring color image data from the color image input device 1, a configuration may be used in which color image data is acquired from an external storage device, a server device, or the like via a network. Further, as the color image output device 3, for example, an image display device such as an organic electroluminescence display or a liquid crystal display, an electrophotographic method or an ink jet method printer that outputs processing results to recording paper or the like is used.

  The present invention may be configured by hardware logic, or may be realized by software using a CPU as follows. That is, the color image processing apparatus 2 develops a CPU (Central Processing Unit) that executes instructions of a control program for realizing each function, a ROM (read only memory) that stores the computer program of the present invention, and various control programs. A random access memory (RAM) and a storage device (recording medium) such as a memory for storing various control programs and various data can be used.

  An object of the present invention is a computer-readable recording in which program code (execution format program, intermediate code program, source program) of a program that performs filter coefficient synthesis processing, which is software that implements the functions described above, is recorded This can also be achieved by supplying the medium to the color image processing apparatus 2 and reading and executing the program code recorded on the recording medium by the computer (or CPU or MPU). As a result, it is possible to provide a portable recording medium on which a computer program code for controlling the filter coefficient synthesis process is recorded.

  This recording medium may be a memory (not shown) such as a ROM, because the processing is performed by a microcomputer, and a program reading device as an external storage device (not shown) is provided in the recording medium. It may be a program medium that can be read by inserting.

  In any case, the stored computer program code may be configured to be accessed and executed by the microprocessor, or the computer program code is read and the read computer program code is illustrated in the microcomputer. The program may be downloaded to a non-program storage area and the computer program code is executed. In this case, it is assumed that the computer program for download is stored in the main device in advance.

  Here, the program medium is a recording medium configured to be separable from the main body, and includes a tape system such as a magnetic tape and a cassette tape, a magnetic disk such as a flexible disk and a hard disk, and a CD-ROM / MO / MD / DVD / Disc systems including optical disks such as CD-R, card systems such as IC cards (including memory cards) / optical cards, mask ROM, EPROM (Erasable Programmable Read Only Memory), EEPROM (Electrically Erasable Programmable Read Only Memory), It may be a medium that carries a fixed computer program including a semiconductor memory such as a flash ROM.

  Further, the color image processing apparatus 2 may be configured to be connectable to a communication network, and may be a medium that fluidly carries the computer program code so as to download the program code via the communication network. When the computer program code is downloaded from the communication network in this way, the computer program for downloading may be stored in the main device in advance or may be installed from another recording medium. The communication network is not particularly limited, and examples thereof include the Internet, intranet, extranet, LAN, CATV communication network, virtual private network, telephone line network, mobile communication network, satellite communication network, and the like. Is available.

  Further, the transmission medium constituting the communication network is not particularly limited. For example, the transmission medium such as IEEE 1394, USB (registered trademark), power line carrier, cable TV line, telephone line, ADSL line, or the like can be used. It can also be used by radio such as infrared, Bluetooth (registered trademark), 802.11 radio, HDR, mobile phone network, satellite line, terrestrial digital network. The present invention can also be realized in the form of a computer data signal embedded in a carrier wave in which the program code is embodied by electronic transmission.

  As described above, the object of the present invention can also be achieved by a program for causing a computer to execute image processing including filter processing for image data for forming an image in which a plurality of pixels are two-dimensionally arranged. it can. This image processing includes a first setting step, a second setting step, a selection step, a conversion step, and a filter processing step. The first setting step is a step of setting a first conversion matrix in accordance with a scaling factor in the vertical direction when the scaling process is performed on the image data. The second setting step is a step of setting a second conversion matrix in accordance with the horizontal scaling factor when the scaling process is performed on the image data. The selection step includes (i) a predetermined image processing condition different from the vertical scaling factor and the horizontal scaling factor, or an image obtained by adding the vertical scaling factor and the horizontal scaling factor to the predetermined image processing condition. This is a step of selecting a reference filter coefficient set based on the processing conditions and (ii) one or more filter coefficient sets stored in advance. In the conversion step, the matrix indicated by the reference filter coefficient set selected in the selection step is linearly converted using the first conversion matrix and the second conversion matrix set in the first setting step and the second setting step. It is a step. The filter processing step is a step of performing filter processing on the image data using the filter coefficient set after the primary conversion in the conversion step. Other applications are the same as those described for the image processing apparatus, and a description thereof will be omitted.

  Another object of the present invention is to perform image processing for filtering image data for forming an image in which a plurality of pixels are arranged two-dimensionally, as exemplified as a processing procedure of a program or a processing procedure of an image processing apparatus. It can also be achieved by the method. This image processing method includes a first setting step, a second setting step, a selection step, a conversion step, and a filter processing step. The first setting step is a step in which the first conversion matrix setting unit sets the first conversion matrix according to the vertical scaling factor when the scaling process is performed on the image data. The second setting step is a step in which the second conversion matrix setting unit sets the second conversion matrix according to the horizontal scaling factor when the scaling process is performed on the image data. In the selection step, the selection unit is configured to select (i) a predetermined image processing condition different from the vertical scaling factor and the horizontal scaling factor, or the vertical scaling factor and the horizontal scaling factor based on the predetermined image processing condition. And (ii) one or a plurality of filter coefficient sets stored in advance, and selecting a reference filter coefficient set. In the conversion step, the coefficient conversion unit converts the matrix indicated by the reference filter coefficient set selected in the selection step into the first conversion matrix and the second conversion matrix set in the first setting step and the second setting step. This is the step of performing the primary conversion. The filter processing step is a step in which the filter processing unit performs a filter process on the image data using the filter coefficient set after the primary conversion in the conversion step. Other applications are the same as those described for the image processing apparatus, and a description thereof will be omitted.

  As described above, the image processing apparatus according to the present invention is an image processing apparatus that performs filter processing on image data for forming an image in which a plurality of pixels are two-dimensionally arranged. A first transformation matrix setting unit that sets a first transformation matrix according to a magnification in the vertical direction when the magnification process is performed on the image data, and a horizontal direction when the image data is subjected to the magnification process A second transformation matrix setting unit for setting a second transformation matrix according to a scaling factor; and (i) predetermined image processing conditions different from the vertical scaling factor and the horizontal scaling factor, or the predetermined A reference filter coefficient set based on the image processing condition obtained by adding the vertical scaling factor and the horizontal scaling factor to the image processing condition, and (ii) one or more filter coefficient sets stored in advance. Selection part to be selected and selected by the selection part A coefficient conversion unit that performs primary conversion on a matrix indicated by a reference filter coefficient set using the first conversion matrix and the second conversion matrix set by the first conversion matrix setting unit and the second conversion matrix setting unit; And a filter processing unit that performs a filter process on the image data using the filter coefficient set subjected to the primary conversion by the coefficient conversion unit. As a result, even when the scaling process performed on the image data is different in the vertical and horizontal scaling ratios, the image quality can be reduced by a single filtering process based on the optimum filter coefficient that can be calculated with a small amount of calculation. Improvements can be realized.

  In the image processing apparatus according to the present invention, the coefficient conversion unit can make the size of the filter coefficient set after the primary conversion different from the size of the reference filter coefficient set, and the filter after the primary conversion The image processing apparatus further includes a size adjustment unit that adjusts a coefficient set to match a size of the reference filter coefficient set and / or a predetermined size, and the filter processing unit uses the filter coefficient set adjusted by the size adjustment unit. It is also possible to configure so as to perform filter processing. Thereby, the reference filter coefficient set of the conventional model originally designed as a necessary size in the image processing apparatus before applying the present invention can be used as it is.

  In the image processing apparatus according to the present invention, the first transformation matrix is a triangular matrix composed of only rational numbers generated by a predetermined calculation using one variable based on the vertical scaling factor. The two transformation matrices may be configured to be a triangular matrix consisting of only rational numbers generated by a predetermined calculation using one variable based on the horizontal scaling factor. As a result, the spatial frequency characteristic of the filter used for the filter processing becomes suitable for a predetermined scaling factor.

  In addition, the image processing apparatus according to the present invention is configured to perform image processing for the conditions included in the predetermined image processing conditions or by adding the vertical scaling factor and the horizontal scaling factor to the predetermined image processing conditions. The conditions included in the conditions may be divided into a plurality of sections, and the first transform matrix, the second transform matrix, and / or the reference filter coefficient set may be switched in each section. Thereby, the filter at the time of filter processing can be finely switched for every condition included in the image processing condition.

  The image processing method according to the present invention is an image processing method for performing filter processing on image data for forming an image in which a plurality of pixels are two-dimensionally arranged. A first setting step for setting a first conversion matrix and a second conversion matrix setting unit according to a vertical scaling factor when performing scaling processing on image data, A second setting step of setting a second transformation matrix in accordance with a horizontal scaling factor when performing the multiplication process; and a selection unit, (i) the vertical scaling factor and the horizontal scaling factor, Are different predetermined image processing conditions, or image processing conditions obtained by adding the vertical scaling factor and the horizontal scaling factor to the predetermined image processing conditions, and (ii) one or more filter coefficients stored in advance And select a reference filter coefficient set based on A selection step, and a coefficient conversion unit, the matrix indicated by the reference filter coefficient set selected in the selection step, the first conversion matrix set in the first setting step and the second setting step, and the first A conversion step of performing a primary conversion using a conversion matrix of 2, and a filter processing step in which a filter processing unit performs a filter process on the image data using the filter coefficient set after the primary conversion in the conversion step. It is characterized by having. As a result, even when the scaling process performed on the image data is different in the vertical and horizontal scaling ratios, the image quality can be reduced by a single filtering process based on the optimum filter coefficient that can be calculated with a small amount of calculation. Improvements can be realized.

  A program according to the present invention is a program for causing a computer to execute image processing including filter processing for image data for forming an image in which a plurality of pixels are two-dimensionally arranged. Includes a first setting step for setting a first transformation matrix according to a vertical scaling factor when the scaling process is performed on the image data, and when the scaling process is performed on the image data. A second setting step for setting a second transformation matrix in accordance with the horizontal scaling factor, and (i) predetermined image processing conditions different from the vertical scaling factor and the horizontal scaling factor, or Based on the image processing condition obtained by adding the vertical scaling factor and the horizontal scaling factor to the predetermined image processing condition, and (ii) one or more filter coefficient sets stored in advance, a reference filter coefficient Set And a matrix indicated by the reference filter coefficient set selected in the selection step, the first transformation matrix and the second transformation set in the first setting step and the second setting step. A conversion step for performing a primary conversion using a matrix; and a filter processing step for performing a filter process on the image data using the filter coefficient set after the primary conversion in the conversion step. is there. As a result, even when the scaling process performed on the image data is different in the vertical and horizontal scaling ratios, the image quality can be reduced by a single filtering process based on the optimum filter coefficient that can be calculated with a small amount of calculation. Improvements can be realized.

DESCRIPTION OF SYMBOLS 1 ... Color image input device, 2 ... Color image processing device, 3 ... Color image output device, 4 ... Operation panel, 20 ... A / D conversion part, 21 ... Shading correction part, 22 ... Input gradation correction part, 23 ... Area separation processing unit, 24 ... color correction unit, 25 ... black generation and under color removal unit, 26 ... spatial filter processing unit, 27 ... scaling unit, 28 ... output gradation correction unit, 29 ... tone reproduction processing unit, 30 ... Filter coefficient setting unit, 100 ... Image forming apparatus, 300 ... Memory (storage unit), 301 ... First transformation matrix setting unit, 302 ... Second transformation matrix setting unit, 303 ... Filter coefficient set selection unit, 304a, 304b ... Coefficient conversion unit, 305a, 305b ... size correction unit, 306a, 306b ... filter coefficient combination unit, 307 ... window function processing unit, 308 ... division coefficient calculation unit, 100 ... image forming apparatus.

Claims (5)

An image processing apparatus that performs filter processing on image data for forming an image in which a plurality of pixels are two-dimensionally arranged,
A first conversion matrix setting unit that sets a first conversion matrix in accordance with a vertical scaling factor when the scaling process is performed on the image data;
A second transformation matrix setting unit for setting a second transformation matrix according to a lateral scaling factor when the scaling process is performed on the image data;
(I) A predetermined image processing condition different from the vertical scaling factor and the horizontal scaling factor, or an image obtained by adding the vertical scaling factor and the horizontal scaling factor to the predetermined image processing condition A selection unit that selects a reference filter coefficient set based on the processing conditions and (ii) one or more filter coefficient sets stored in advance;
The matrix indicated by the reference filter coefficient set selected by the selection unit is converted into a first order using the first conversion matrix and the second conversion matrix set by the first conversion matrix setting unit and the second conversion matrix setting unit. A coefficient conversion unit for conversion;
A filter processing unit that performs a filter process on the image data using the filter coefficient set after the primary conversion by the coefficient conversion unit;
An image processing apparatus comprising: It is possible to make the size of the filter coefficient set after the primary conversion different from the size of the reference filter coefficient set by the coefficient conversion unit,
A size adjusting unit that adjusts the filter coefficient set after the primary conversion to match the size of the reference filter coefficient set and / or a predetermined size;
The image processing apparatus according to claim 1, wherein the filter processing unit performs a filter process using the filter coefficient set adjusted by the size adjusting unit. The first transformation matrix is a triangular matrix consisting of only rational numbers generated by a predetermined calculation using one variable based on the scaling factor in the vertical direction,
3. The image according to claim 1, wherein the second transformation matrix is a triangular matrix including only rational numbers generated by a predetermined calculation using one variable based on the scaling factor in the horizontal direction. Processing equipment.   A plurality of categories for each condition included in the predetermined image processing condition, or for each condition included in the image processing condition obtained by adding the vertical scaling factor and the horizontal scaling factor to the predetermined image processing condition. The image processing according to claim 1, wherein the first transformation matrix, the second transformation matrix, and / or the reference filter coefficient set are switched in each division. apparatus. An image processing method for performing filter processing on image data for forming an image in which a plurality of pixels are two-dimensionally arranged,
A first setting step in which a first conversion matrix setting unit sets a first conversion matrix in accordance with a vertical scaling factor when performing a scaling process on the image data;
A second setting step in which a second conversion matrix setting unit sets a second conversion matrix according to a horizontal scaling factor when the scaling process is performed on the image data;
(I) a predetermined image processing condition different from the vertical scaling factor and the horizontal scaling factor, or the vertical scaling factor and the horizontal scaling factor in the predetermined image processing condition; A selection step of selecting a reference filter coefficient set based on (ii) one or more pre-stored filter coefficient sets;
The coefficient conversion unit represents the matrix indicated by the reference filter coefficient set selected in the selection step as the first conversion matrix and the second conversion matrix set in the first setting step and the second setting step. A conversion step for primary conversion using,
A filter processing unit that performs a filter process on the image data using the filter coefficient set after the primary conversion in the conversion step;
An image processing method comprising:

Images