JP2009051219A - Printing apparatus, image data file processor, selection method of image data file, and computer program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To determine the fuzziness degree of the image of an image data file with little load in the image data file which holds the image in the form of the coefficient of a discrete frequency component as an object. <P>SOLUTION: Image data including a plurality of coefficients representing the pattern of the change of the color of the image are acquired (S10). A plurality of basic pattern data including a plurality of basic coefficients representing the basic pattern of the change of the color are prepared. Additionally, one selected basic pattern data is selected from the plurality of the basic pattern data based on the comparison of both (S20). Thereafter, the fuzziness degree of the image of the image data is calculated from the basic fuzziness degree corresponding to the selected basic pattern data (S30 to S70). <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a technique for determining the degree of blur of an image of an image data file.

  Conventionally, there is a technique for selecting an in-focus image from a plurality of photographic images. For example, in a certain prior art, when taking a photograph, the image sensor is moved from the reference position to two positions before and after, and an evaluation function is calculated for the image obtained at each position. Then, the reference position of the image sensor is changed toward the better one of the evaluation function values before and after. Such processing is repeated to determine the focus position of the image sensor.

The evaluation function is calculated as follows. That is, frequency conversion is performed on an 8 pixel × 8 pixel block in an image to obtain 8 × 8 frequency components a _kl (k = 1 to 8, l = 1 to 8). Among the “8 × 8 frequency components a _kl (k = 1 to 8, l = 1 to 8) of each block, the sum of absolute values of the high frequency components (k + 1> 9) is the low frequency component (k + 1 = 7). The sum of “value divided by the sum of absolute values of” is the evaluation function. There exists patent document 1 as literature regarding such a technique.

Japanese Patent Laid-Open No. 4-170872

  Further, in a certain prior art, the following processing is performed when determining whether or not an image is blurred. That is, the square root g (x, y) of the square sum of the brightness differences fx and fy between the target pixel in the image and the pixels adjacent to the target pixel in the X and Y directions is calculated. Then, the average value of g (x, y) for all the pixels in the image is calculated as “sharpness SL”. It is determined whether the image is blurred based on the size of the SL. There exists patent document 2 as a literature regarding such a technique.

Japanese Patent Laid-Open No. 10-340332

However, the former technique has 28 components that are high frequency components (k + 1> 9) out of 8 × 8 frequency components a _kl (k = 1 to 8, l = 1 to 8) and low frequency components ( It is necessary to calculate a total of 48 components including 21 components with k + 1 <7) for all blocks of each image. That is, the calculation amount is enormous and the processing load is large. Further, this technique is a technique that should be applied to comparison between images having the same composition.

  On the other hand, the latter technique does not hold an image as color information for each pixel as in the BMP format, but stores an image in the form of discrete frequency component coefficients. It cannot be applied directly. That is, when the image data is JPEG format data, the entire image must be once converted into a format (for example, BMP format) having color information for each pixel. Therefore, the processing load increases.

  The present invention has been made in order to deal with the above-described problems, and is intended for image data files that hold images in the form of discrete frequency component coefficients. Degree can be determined.

  In order to achieve the above object, the present invention performs the following processing when calculating the blur level of an image of image data. The “blurring degree” is a parameter that represents how much the image is blurred, and has a larger value as the image becomes more blurred.

  The image data is image data including a plurality of coefficients representing a pattern of color change of an image along a predetermined direction. A plurality of coefficients included in the image data are coefficients corresponding to different frequency components. The color change pattern of the image can be a luminance (lightness) change pattern. Also, the color change pattern of the image may be a pattern of change in the value of each color component of RGB or YCrCb.

  When calculating the degree of blur of the image of the image data, first, the image data includes a plurality of coefficients representing a pattern of color change of the image along a predetermined direction, and each of the plurality of coefficients has a different frequency component. The image data that is the coefficient corresponding to is acquired. Also, a plurality of basic pattern data each including a plurality of basic coefficients representing a basic pattern of color change is prepared. The plurality of basic coefficients are coefficients corresponding to different frequency components. The basic pattern data includes data representing a basic blur level that represents a blur level at a color boundary.

  Then, based on the comparison between the plurality of coefficients included in the image data and the plurality of basic coefficients respectively included in the plurality of basic pattern data, one selected basic pattern data is selected from the plurality of basic pattern data. . The “comparison” between the plurality of coefficients included in the image data and the plurality of basic counts includes direct comparison between the plurality of coefficients included in the image data and the plurality of basic counts, as well as a predetermined number of coefficients. Indirect comparison performed between a plurality of numbers obtained by performing the above-described processing and a plurality of basic counts is also included. Thereafter, based on the basic blur level associated with the selected basic pattern data, the blur level of the image of the image data is calculated.

  According to such an aspect, it is possible to determine the blur level of the image of the image data file with a small load for an image data file that holds an image in the form of discrete frequency component coefficients.

  The present invention can also be configured as a printing apparatus that selects and prints image data candidates from a plurality of image data. The printing apparatus includes a data acquisition unit, a basic pattern memory, a basic pattern selection unit, a candidate determination unit, and a printing unit.

  The data acquisition unit can acquire one image data from a plurality of image data. The basic pattern memory can store a plurality of basic pattern data each including a plurality of basic coefficients representing a basic pattern of color change. The plurality of basic coefficients are coefficients corresponding to different frequency components. The basic pattern data includes data representing the basic blur level indicating the blur level at the color boundary portion.

  The basic pattern selection unit, based on the comparison between the plurality of coefficients included in the acquired image data and the plurality of basic coefficients included in each of the plurality of basic pattern data stored in the basic pattern memory, One selected basic pattern data is selected from the pattern data. The candidate determining unit calculates the blur level of the image data based on the basic blur level associated with the selected basic pattern data, and determines whether or not the image data acquired based on the blur level is a candidate. The printing unit prints the image data selected as a candidate.

  According to such an aspect, it is possible to select and print image data candidates with a small load from a plurality of image data holding images in the form of discrete frequency component coefficients.

  Note that the basic pattern selection unit displays the pattern of the color change of the image along the predetermined direction in the target area, which is a partial area of the image of the image data, when the predetermined first condition is satisfied. The selected basic pattern data can be selected based on a comparison between the plurality of coefficients to be represented and the plurality of basic coefficients respectively included in the plurality of basic pattern data stored in the basic pattern memory.

  In addition, the candidate determining unit can determine the partial blur level of the target region based on the basic blur level of the selected basic pattern data. Then, the candidate determination unit can determine whether or not the acquired image data is a candidate based on the partial blurring levels of a plurality of target regions included in the acquired image data image. With such an aspect, it is possible to calculate the blurring degree of the entire image in consideration of the partial blurring degree of each part in the image.

  Instead of selecting the selected basic pattern data for each target area in the image, the selected basic pattern data may be selected by comparing the entire image with each basic pattern.

  The basic coefficient is preferably a coefficient corresponding to the AC component of the frequency components. Then, the basic pattern selection unit selects the selected basic pattern data based on the comparison between the coefficient corresponding to the AC component among the plurality of coefficients in the target region and the plurality of basic coefficients included in each of the plurality of basic pattern data. It is preferable to select.

  With such an aspect, it is possible to compare the color pattern with the basic pattern without being influenced by the absolute shade of the color or the degree of light and dark. For this reason, the number of basic pattern data to be prepared for comparison can be reduced.

  Note that the plurality of basic coefficients are all positive numbers and are preferably coefficients whose sum is a constant value. And it is preferable that the basic pattern selection unit further calculates a coefficient corresponding to each frequency component for the target region based on a ratio of magnitudes of absolute values of the frequency components so that the total is a constant number. .

  With such an aspect, the color pattern and the basic pattern can be compared without being influenced by the absolute magnitude of the change width of the color (including luminance). For this reason, the number of basic pattern data to be prepared for comparison can be reduced.

  Note that the plurality of basic pattern data preferably has data in a predetermined order. And it is preferable that a basic pattern selection part performs the following processes. That is, when selecting the selected basic pattern data, the plurality of basic coefficients of the plurality of basic pattern data are sequentially compared with the plurality of coefficients in the target region according to the order. Then, the first basic pattern in which the difference between the plurality of basic coefficients and the plurality of coefficients is smaller than a predetermined reference is selected as selected basic pattern data.

  According to such an embodiment, the selected basic pattern data having a plurality of basic coefficients that approximates to some extent to a plurality of coefficients in the target area without comparing the basic coefficients of all the basic pattern data and the coefficients of the target area. You can choose.

  In addition, the plurality of basic coefficients are preferably coefficients representing a luminance change pattern in which the luminance changes from the first value to the second value within a predetermined section as a basic pattern of color change. The basic blur level is preferably a value corresponding to the width of a section where the luminance changes from the first value to the second value within a predetermined section.

  Then, in the first selected basic pattern data selected based on the first target area, a section where the luminance changes, and a second target area adjacent to the first target area side by side in a predetermined direction In the second selected basic pattern data selected on the basis, the section in which the luminance changes is located at the end of the other target area in the predetermined section of the target area, and the direction of the change in the luminance is If they are the same, the candidate determination unit preferably performs the following processing.

  That is, the candidate determining unit calculates the partial blurring degree of at least one of the first and second target regions based on the sum of the basic blurring degrees of the first and second selected basic pattern data. With such an aspect, the partial blurring degree can be calculated even when the boundary between regions having different colors in the image is blurred across the target region.

  The basic pattern selection unit further determines the partial blurring degree of at least one of the first and second target areas and the basic blurring degree of the selected basic pattern data of the first and second target areas even in the following cases. It is preferable to calculate based on the sum of. In this case, in the first and second selected basic pattern data, the section where the luminance changes exists from the end on the other target area side of the predetermined section to a position inside by a width equal to or smaller than the predetermined value. And the direction of change in luminance is the same.

  In image processing performed in advance on image data, a constant luminance portion that should not occur originally may occur at the end of the target area. However, if it is set as the above aspect, a partial blurring degree can be calculated appropriately also about such an object area | region.

  Note that the predetermined sections of the plurality of basic patterns may have the same size. In such an aspect, it is not necessary to consider the size of the area of the basic pattern in the comparison between the coefficient of the target area and the basic coefficient of the basic pattern data. For this reason, it is possible to easily compare the coefficient of the target region with the basic coefficient of the basic pattern data.

  Note that the size of the predetermined section of the target region is preferably 8 pixels of the image data. The DCT coefficient included in the image data in the JPEG format is a coefficient related to a target area of 8 pixels × 8 pixels. For this reason, if it is set as the above aspects, it is possible to easily compare the color change pattern of the target region with the pattern of the basic pattern data.

  The plurality of basic patterns may include basic patterns having different predetermined section sizes. With such an aspect, it is possible to select an approximate basic pattern from a plurality of basic patterns for various color change patterns having different sizes (number of pixels).

  Moreover, it is preferable that the first condition is that a sum of absolute values of a plurality of coefficients in the target region is larger than a predetermined reference value. With such an aspect, it is possible to omit the process of selecting the selected basic pattern data for an area with little color change. For this reason, it is possible to reduce the processing load for calculating the degree of blur of the image.

  Note that the basic pattern selection unit is further configured based on a difference in luminance between two adjacent target regions when the first condition is not satisfied and the predetermined second condition is satisfied. It is preferable to select the selected basic pattern data. More specifically, when the luminance difference exceeds a predetermined value, it is preferable to select the basic pattern data having the smallest basic blur level from the basic pattern data as the selected basic pattern data.

  With such an aspect, even when the color boundary overlaps the boundary of the target region, the selected basic pattern data can be appropriately associated with the color boundary. Note that the second condition is that the absolute value of a plurality of coefficients is smaller than a predetermined reference value for each of two adjacent target areas along a predetermined direction related to the color change pattern. It is preferable to do.

  In addition, the candidate determination unit preferably performs the following processing when determining the degree of partial blur. That is, the selected basic pattern data is selected based on a plurality of coefficients representing the color change pattern of the image of the target area along the horizontal direction of the image as one of the predetermined directions related to the color change pattern. Then, the horizontal blurring degree of the target area is calculated based on the basic blurring degree of the selected basic pattern data. On the other hand, the selected basic pattern data is selected based on a plurality of coefficients representing the color change pattern of the image of the target area along the vertical direction of the image as one of the predetermined directions related to the color change pattern. Then, the vertical blurring degree of the target area is calculated based on the basic blurring degree of the selected basic pattern data.

  Thereafter, the candidate determination unit determines the partial blurring degree of the target region based on the horizontal blurring degree and the vertical blurring degree. With such an aspect, it is possible to determine the blur level of an image in consideration of the blur levels in two directions that are 90 degrees to each other.

  It is preferable that the candidate determination unit determines the partial blur level of the target region based on the larger one of the horizontal blur level and the vertical blur level. According to such an aspect, even when the image is blurred in a certain direction, or when one of the horizontal blur level and the vertical blur level is erroneously calculated to a small value for some reason, the blur level of the image Is likely to be able to be determined appropriately.

  Note that the candidate determining unit determines a target area that is a predetermined size included in the image of the acquired image data and has a partial blurring degree smaller than the first reference value than the second reference value. It is preferable to use the acquired image data as a candidate when there is a region including many.

  With such an aspect, if there is a region with a clear outline in the image of the image data, the image data can be a candidate for printing even if the contour is not clear in other regions. The region having a predetermined size is preferably a region having a length of 0.4 to 1.6 cm and a width of 0.4 to 1.6 cm when an image is printed on the L plate.

  The first and second reference values are preferably determined based on the number of pixels of the image data image and the size of the image data image when printing. This is because the result of determination by a human as to whether or not a certain image is blurred may differ depending on the size of one pixel of image data that is expressed during printing.

  A plurality of basic pattern data may be classified into a plurality of basic pattern data groups each including one or more basic pattern data. In such an embodiment, the basic pattern selection unit selects one of the plurality of basic pattern data groups based on a part of the plurality of coefficients of the image data when selecting the selected basic pattern data. It is preferable to include a group selection unit. Then, the basic pattern selection unit is configured to compare the plurality of basic pattern data based on the comparison between the plurality of coefficients included in the image data and the plurality of basic coefficients included in the basic pattern data of the selected basic pattern data group. It is preferable to select one selected basic pattern data from among them.

  With such an embodiment, the number of basic patterns for which the coefficients are compared in advance can be limited. Therefore, the processing load can be reduced as compared with the aspect in which the coefficients are compared for all the basic patterns.

  In the above aspect, the following treatment is preferably performed. That is, the basic pattern selection unit displays the pattern of the color change of the image along the predetermined direction in the target area, which is a partial area of the image of the image data, when the predetermined first condition is satisfied. The selected basic pattern data is selected based on a comparison between the plurality of coefficients to be represented and the plurality of basic coefficients respectively included in the plurality of basic pattern data stored in the basic pattern memory. Then, the candidate determining unit determines the partial blurring degree of the target region based on the basic blurring degree of the selected basic pattern data. Then, the candidate determination unit determines whether or not the acquired image data is a candidate based on the partial blurring degrees of the plurality of target regions included in the image of the acquired image data. Further, the basic coefficient can be a coefficient corresponding to the AC component of the frequency components. Then, the basic pattern selection unit selects the selected basic pattern data based on the comparison between the coefficient corresponding to the AC component among the plurality of coefficients in the target region and the plurality of basic coefficients included in each of the plurality of basic pattern data. select.

  Furthermore, the plurality of basic coefficients may be coefficients that have a constant absolute value. In such an aspect, the basic pattern selection unit further calculates a coefficient corresponding to each frequency component so that the sum of absolute values becomes a constant value based on each frequency component for the target region. Is preferred.

  In such an aspect, with respect to the luminance change patterns having different absolute magnitudes, the selected basic pattern data is selected from the basic pattern data limited to a certain number regardless of the magnitude of the luminance change. Can do. Then, it is possible to select the selected basic pattern data by distinguishing between a target region having a positive coefficient corresponding to the AC component and a target region having a negative coefficient.

  The group selection unit selects one of the plurality of basic pattern data groups based on the sign of the coefficient corresponding to the first and second low frequency AC components among the plurality of coefficients. It is preferable to select.

  With such an aspect, it is possible to select the basic pattern data group with less load compared to the aspect of selecting the basic pattern data group by calculating the color density at each pixel position in the target region.

  In addition, the basic pattern selection unit, when a predetermined first condition is satisfied, displays an image color change pattern along a predetermined direction in a target area that is a partial area of the image of the image data. It is preferable to select the selected basic pattern data based on a comparison between a plurality of coefficients to be represented and a plurality of basic coefficients respectively included in the plurality of basic pattern data stored in the basic pattern memory.

  And it is preferable that a candidate determination part performs the following processes. In other words, the candidate determining unit determines the partial blurring degree of the target region included in the image of the acquired image data based on the basic blurring degree of the selected basic pattern data. The candidate determining unit determines a first evaluation value corresponding to the difference between the maximum value and the minimum value of the color density in the target region based on the plurality of coefficients of the target region. Then, the candidate determining unit determines the second evaluation values related to the plurality of target regions included in the acquired image data based on the first evaluation values of the target regions included in the plurality of target regions. To do. Thereafter, when the second evaluation value satisfies the predetermined second condition, the candidate determining unit selects the acquired image data based on the partial blurring degree of each target region included in the plurality of target regions. Decide whether or not.

  With such an aspect, it is possible to determine whether or not to determine the partial blurring degree of each target area based on the partial blurring degree of the plurality of target areas in consideration of the difference in color density in the image. it can. Therefore, it is possible to reduce the possibility of making a determination regarding image data based on the partial blurring degree of a plurality of target regions determined by treating a portion that is not a boundary between different color regions as a boundary.

The second condition can be performed based on, for example, a comparison between the second evaluation value and a predetermined threshold value. If the second evaluation value is the same under other conditions, and the evaluation value becomes larger as the first evaluation value is larger, “the second evaluation value is a predetermined threshold value”. “A value greater than the value” can be the second condition. In addition, if the second evaluation value is the same under other conditions, if the second evaluation value is an evaluation value that becomes smaller as the first evaluation value is larger, “the second evaluation value is a predetermined value. “A value smaller than the threshold value” can be set as the second condition.
That is, the second condition may be any determination condition that uses the second evaluation value when determining whether the condition is satisfied.

  The plurality of target areas are a plurality of target areas that exist continuously along a predetermined direction, and can be a plurality of target areas that satisfy a predetermined condition regarding the selected basic pattern data.

  According to such an aspect, a portion that spans a plurality of target regions in an image and that changes in color density along one direction takes into account the amount of change in color density of that portion, A decision regarding blurring can be made. That is, whether or not to determine the partial blurring degree of each target area included in the part based on the partial blurring degree of a plurality of target areas in consideration of the amount of change in the color density of the part. Can do. Therefore, it is possible to reduce the possibility of making a decision regarding image data by treating a portion that is not a boundary between different color areas as a boundary between different color areas. Note that the second evaluation value can be, for example, the total value of the first evaluation values of the target regions included in the plurality of target regions.

  The predetermined condition relating to the selected basic pattern data preferably includes at least a part of the following condition. The condition is a condition that the basic pattern represented by the selected basic pattern data includes a portion whose luminance changes at at least one end of the basic pattern.

  With such an aspect, it is possible to accurately determine whether a certain target region is a target region included in a portion where the color density continuously changes.

“Condition A includes condition a1 as at least a part” includes “a condition a1 is included as a condition (and condition) that must be satisfied in order to satisfy condition A” and “condition A In order to satisfy the condition, both the condition a1 and the condition that the condition a1 is included as the condition (or condition) that at least one of the other one or more conditions should be satisfied are included.
Furthermore, an aspect in which the condition A is configured by a combination of lower-order conditions (AND condition and OR condition) is also included. That is, “Condition A includes at least part of condition a1” means that “in order to determine whether condition A is satisfied for all objects, it is necessary to determine whether condition a1 is satisfied. Represents the relationship.

  The candidate determination unit is a basic coefficient of the selected basic pattern data that corresponds to the alternating current component of the frequency component and is the sum of the absolute values of the coefficients corresponding to the alternating current component among the frequency components. It is preferable to obtain the first evaluation value based on the ratio of the coefficient to the sum of absolute values.

  According to such an aspect, the color density of each pixel position is calculated, and compared with the aspect of calculating the first evaluation value corresponding to the difference between the maximum value and the minimum value of the color density in the target region, The first evaluation value can be calculated with a small load. It is preferable to obtain the first evaluation value by further dividing the above ratio by 2√2.

  The present invention can be realized in various forms, for example, a printing method and a printing apparatus, a printing control method and a printing control apparatus, a method for determining whether an image of photographic image data is blurred, and In the form of an apparatus for implementing the method, a computer program for realizing the function of the method or apparatus, a recording medium recording the computer program, a data signal including the computer program and embodied in a carrier wave, etc. Can be realized.

A. First embodiment:
FIG. 1 is a perspective view showing a printer 1 according to a first embodiment of the present invention. The printer 1 is a printer that can perform printing based on an image data file stored alone in a storage medium without being connected to an external computer. The printer 1 includes a print head (not shown) that performs printing by ejecting ink droplets, an auto seed feeder 20 that supplies printing paper, and a paper discharge tray 30 that receives printing paper on which an image is printed. , A liquid crystal display 40, a button group 50 for performing various operations, a card slot 70 for inserting a memory card and reading data, a CPU 100, a main memory 150, and a ROM 160. In FIG. 1, the CPU 100, the main memory 150, and the ROM 160 are displayed outside the printer 1 for easy explanation.

  In the card slot 70, a memory card MC such as a compact flash (registered trademark) card, an SD card, a miniSD card, a memory stick, a smart media card or the like can be directly inserted into the slot, or can be inserted via an adapter (see FIG. 1 arrow A1). Then, the CPU 100 can acquire a plurality of image data files stored in those memory cards MC via the card slot 70.

  FIG. 2 is an enlarged plan view showing an operation panel provided with the liquid crystal display 40 and the button group 50. In the center of the operation panel, a 2.5 inch size liquid crystal display 40 is provided. The liquid crystal display 40 can display a color image. The liquid crystal display 40 can display an image of an image data file stored in the memory card MC, for example. The liquid crystal display 40 can also display, for example, information indicating the state of the printer 1 and an operation menu used when a user inputs an instruction to the printer 1 via the button group 50.

  A function unit of the CPU 100 that displays a user interface screen on the liquid crystal display 40 to prompt the user to input an instruction and receives an instruction from the user via the button group 50 is shown as a user interface unit 105 in FIG. .

  Buttons 52 to 62 for performing various operations are provided on both sides of the liquid crystal display 40. These are collectively referred to as a button group 50. The power button 52 is a button for turning on / off the power of the printer 1. The mode button group 54 is a plurality of buttons for the user to switch the operation mode of the printer. One button corresponds to one operation mode. The setting button 56 is a button that is pressed when performing various maintenance of the printer 1 and operation setting of each mode.

  The menu button group 58 is a plurality of buttons for operating various numerical values and icons displayed on the liquid crystal display 40. The start button group 60 is two buttons 601 and 602 for selecting whether to perform color printing or monochrome printing. One button corresponds to each of color printing and monochrome printing. The stop button 62 is a button for stopping the operation being executed by the printer 1.

  The menu button group 58 includes an OK button 582 for inputting that the setting contents presented on the liquid crystal display 40 are accepted, and a return button for displaying the contents displayed on the liquid crystal display 40 one step before. 584, a selection button group 586 including four buttons for selecting one of a plurality of icons and menus displayed on the liquid crystal display 40, and a number of print sheets to be printed. A print number button group 588. By providing such a group of buttons, the user can easily select and print an image data file stored in the memory card even in the printer 1 equipped with only a relatively small display.

  2 is inserted into the card slot 70 of the printer 1 (see arrow A1 in FIG. 1), and the “memory card” in the mode button group 54 is inserted. The display when the "" button is pressed is shown. In the upper part of the liquid crystal display 40, three types of menus “print all” M11, “select print” M12, and “automatic print” M13 are displayed. The “automatic printing” menu M13 is surrounded by a frame S1. The state in FIG. 2 is a state in which “automatic printing” M13 is provisionally selected.

  In the lower part of the liquid crystal display 40, a display M21 for explaining the processing content of “automatic printing” which is provisionally selected at this time is displayed. “Random printing” is a process in which a printing apparatus automatically selects and prints an image data file from a plurality of image data files stored in a memory card. In the state shown in FIG. 2 on the liquid crystal display 40, the user operates the selection button group 586 (see FIG. 2) to confirm the selection of “Random Print” and to perform “Random Print” on the printer 1. Can be executed.

  FIG. 3 is a flowchart showing the processing content of “automatic printing”. FIG. 3 shows processing performed by the CPU 100 after the user presses the color printing button 601 or the monochrome printing button 602 in the state where the display of FIG. 2 is displayed on the liquid crystal display 40.

  In step S10, the CPU 100 reads one image data file from the memory card MC to the main memory 150. The image data file read from the memory card MC is a JPEG format image data file. A functional unit of the CPU 100 that performs the function of step S10 is shown as a data acquisition unit 110 in FIG.

  FIG. 4 is a diagram illustrating DCT coefficients for an image of a target area stored with image data in JPEG format. The image data in JPEG format is a coefficient (DCT coefficient) F00 representing frequency components of a plurality of discrete frequencies with respect to the image of each target area which is a small area of 8 pixels × 8 pixels obtained by dividing the image area. ~ F77 are stored. This DCT coefficient can be obtained by performing a discrete cosine transform (DCT) on data representing the image of the target region with the luminance of each pixel. A total of 64 DCT coefficients of [8 rows for vertical frequency] × [8 columns for horizontal frequency] are stored for each target region.

  Note that monochrome image data in the JPEG format retains the above DCT coefficients for luminance. Also, the color image data in the JPEG format holds the DCT coefficients of the Y component (luminance), Cr component (red-green component), and Cb component (blue-yellow component) of the YCrCb color system.

  In step S20 of FIG. 3, edge pattern matching is performed. Note that the processing in steps S20 and S30 is performed in the horizontal and vertical directions of the image, respectively. In the present embodiment, first, the processes in steps S20 and S30 are performed in the horizontal direction of the image, and then the processes in steps S20 and S30 are performed in the vertical direction of the image through the determination in step S40.

  In the edge pattern matching in step S20, a corresponding basic pattern is selected from a plurality of basic patterns and associated with each target region of 8 pixels × 8 pixels of the image in the image data file. At that time, the direction of luminance change is also determined. Further, for each target region, the horizontal partial blur level Bbh and the vertical partial blur level Bbv are calculated. The edge pattern matching will be described in detail later. A functional unit of the CPU 100 that performs the edge pattern matching function in step S20 is shown as a basic pattern selection unit 120 in FIG.

  In step S30, edge connection processing is performed. In the edge linking process, the following processing is performed on the boundary where an area of a certain color and another color area contact in the image, that is, the end (edge) of an object having a certain color in the image. .

  First, in a horizontal direction or a vertical direction, a place (target region) where a color changes from one color to another color through a plurality of target regions is specified. The place where a color changes from one color to another color is a boundary where a certain color region and another color region contact each other. Therefore, such a boundary is blurred beyond the width of one target area (equivalent to 8 pixels). After such target regions are specified, the horizontal partial blur level Bbh and the vertical partial blur level Bbv of those target regions are modified. The edge connection process will be described in detail later.

  In step S40, it is determined whether or not the processing in steps S20 and S30 has been performed for both the horizontal direction and the vertical direction of the image. If the processes in steps S20 and S30 have not been performed for the vertical direction yet, the process returns to step S20, and the processes in steps S20 and S30 are performed for the vertical direction. If the processes in steps S20 and S30 have been performed for both the horizontal direction and the vertical direction of the image, the process proceeds to step S50.

  In step S50, for each target region, the larger of the horizontal partial blur level Bbh and the vertical partial blur level Bbv of the target region is set as the partial blur level Bb of the target region.

  In step S60, it is determined whether or not each target area is in focus, that is, whether or not the target area is in focus. In step S60, it is determined that the target area whose partial blur level Bb is smaller than the predetermined threshold value Thbb is the “target area in focus”.

  Hereinafter, the target area determined to be in focus is referred to as a “focus area”. On the other hand, the target area whose partial blurring degree Bb is equal to or greater than the predetermined threshold Thbb is determined to be a “target area that is not in focus”. Hereinafter, a target area determined to be out of focus is referred to as an “out-of-focus area”.

  As a classification of the target area, there is a “flat area” in addition to the “in-focus area” and the “in-focus area”. Whether each target region is a “flat region” is determined in edge pattern matching (step S20) and edge connection processing (step S30). For example, when there is no color-to-color boundary in a certain target area, the target area is a “flat area”. The “flat region” will be described later.

  In step S70, the blur level Rf of the image is calculated based on the number of in-focus areas in the image. Specifically, as shown in the following formula (1), a ratio Rf obtained by dividing the number Nf of in-focus areas in an image by the sum of the number Nf of in-focus areas and the number Nn of out-of-focus areas. Calculate At that time, the flat region is not considered. This in-focus area ratio Rf is the degree of blur of the image of the image data file.

  Rf = Nf / (Nf + Nn) (1)

  In step S80, it is determined whether or not the image is in focus, that is, whether or not the image is “in focus”. In step S80, when the blur level Rf of the image is greater than or equal to a predetermined threshold value Thrf, it is determined that the image as a whole is “in focus”. Hereinafter, an image determined to be in focus is referred to as a “focused image”. On the other hand, an image that is not determined to be in focus is referred to as an “out-of-focus image”. A functional unit of the CPU 100 that performs the functions of steps S30 to S80 is shown as a candidate determination unit 130 in FIG.

  In step S90, the next process to be performed is determined according to whether or not the image of the image data file currently being processed is a focused image. If the image in the image data file is a focused image, the image in the image data file is printed in step S100. On the other hand, when the image of the image data file is an out-of-focus image, the process of step S100 is skipped. The functional unit of the CPU 100 that performs the function of step S100 is shown in FIG.

  In step S110, it is determined whether or not the processing in steps S10 to S90 has been performed for all the image data files in the memory card. If there is an image data file in the memory card that has not been subjected to the processes in steps S10 to S90, the process returns to step S10 again. In step S10, one image data file is read from the image data files that have not been subjected to the processes in steps S10 to S90. On the other hand, when the processing of steps S10 to S90 has been performed for all the image data files in the memory card, the processing for automatic printing ends.

  By performing the processing of FIG. 3 described above, the printer 1 can automatically select and print only the image data file of the focused image from the image data file of the memory card.

  FIG. 5 is a flowchart showing the processing contents of edge pattern matching (see step S20 in FIG. 3). Here, a case where edge pattern matching is performed in the horizontal direction will be described as an example. Similar processing is performed when edge pattern matching is performed in the vertical direction.

  First, the DCT coefficient of the image data file in JPEG format will be further described with reference to FIG. The DCT coefficient F00 corresponds to the average value of the luminance of the target area (8 pixels × 8 pixels). The coefficient F00 may be referred to as “DC component”.

  On the other hand, the other 63 DCT coefficients F01 to F77 may be referred to as “AC components”. These DCT coefficients F01 to F77 represent the respective frequency components when the luminance density pattern of the pixel in the target region is approximated by superimposing the cosine waves of the respective frequencies.

  In the figure, the seven DCT coefficients F01 to F07 arranged in the horizontal direction are obtained by superimposing cosine waves of different frequencies on a horizontal luminance shading pattern obtained by averaging the luminances of the pixels in the target region in the vertical direction. Represents each frequency component when approximated by combination. In FIG. 4, the coefficient on the right represents a high frequency component, and the coefficient on the left represents a low frequency component.

  On the other hand, the seven DCT coefficients F10 to F70 arranged in the vertical direction in the figure represent cosine waves having different frequencies from each other in the vertical luminance gradation pattern obtained by averaging the luminances of the pixels in the target region in the horizontal direction. Each frequency component is expressed when approximated by superposition of. In FIG. 4, the lower coefficient represents the higher frequency component, and the upper coefficient represents the lower frequency component.

  In step S310 of FIG. 5, eight DCT coefficients F00 to F07 in the horizontal direction of a certain target region are acquired. In step S320, the sum Sfh of absolute values of the DCT coefficients F01 to F07 of the AC components is calculated.

  In step S330, it is determined whether Sfh is greater than a predetermined threshold value Thsf. If Sfh is less than or equal to threshold value Thsf, the process proceeds to step S510 (see A in FIG. 5). The processing after step S510 will be described later. On the other hand, when Sfh is larger than the threshold value Thsf, the process proceeds to step S340. The case where Sfh is larger than the threshold Thsf means that there is a certain level of luminance fluctuation in the horizontal direction within the target region.

  In step S340, normalization coefficients Fr01 to Fr07 are calculated based on the DCT coefficients F01 to F07 of the AC component in the horizontal direction of the target region. Fr01 to Fr07 are obtained by normalizing the absolute values of the DCT coefficients F01 to F07 according to the following equation (3). Expression (3) is an expression for calculating the normalization coefficients Fr01 to Fr07 corresponding to the DCT coefficients F01 to F07, respectively, so that the sum is 1 based on the ratio of the absolute values of the DCT coefficients F01 to F07. It is. In addition, i is an integer of 1-7.

  In this way, the absolute value is calculated for a plurality of coefficients, and according to the ratio of the magnitudes of the absolute values, the same number of new coefficients as the original coefficient that is 1 is calculated. In this specification, it is called “normalization”.

  By the way, in the two luminance change patterns in which the change in luminance is vertically reversed, the absolute values of the corresponding DCT coefficients are the same and the signs are opposite. For example, there are a first pattern in which there is a portion with constant luminance at the left end and the luminance increases from there to the right, and a pattern obtained by inverting the first pattern up and down. This is a combination with the second pattern in which the brightness decreases from there to the right.

  In the present embodiment, the absolute values of the DCT coefficients F01 to F07 are normalized as described above (see step S340 in FIG. 5), and then matching with the basic pattern is performed. For this reason, it is possible to perform matching between two patterns in which the change in luminance is inverted up and down with one basic pattern. That is, according to the present embodiment, the number of basic patterns to be prepared for matching is different from a mode in which basic patterns corresponding to two patterns whose brightness changes are vertically inverted from each other. Can be reduced.

  In step S350, the target region normalization coefficient Fr0i (i = 1 to 7) and the luminance change basic pattern Pj (j is 01 to 07, 11 to 15, 21 to 23) according to the following equation (4). , 31), the sum SDj of the absolute values of the differences of the normalized DCT coefficients Cfnji.

  FIG. 6 is a diagram showing the luminance level L1i of the basic pattern P01, the one-dimensional DCT coefficient Cf1i, and the normalized DCT coefficient Cfn1i (i = 1 to 7). As shown in the upper part of FIG. 6, the basic pattern P01 has a luminance level of 1 at the pixel position 0, a luminance level of the pixel positions 1 to 7 is 0, and the luminance linearly from the pixel position 0 to 1. This is a luminance change pattern in which the level decreases. This basic pattern P01 is expressed by eight parameters of discrete luminance levels L1i (i = 0 to 7) shown in the lower table of FIG.

  The one-dimensional DCT coefficients obtained by performing the one-dimensional DCT on the luminance level L1i (i = 0 to 7) are the eight parameters Cf1i (i = 0 to 7) shown in the second row of the table. Further, out of the eight one-dimensional DCT coefficients Cf1i (i = 0 to 7), the absolute values of the seven one-dimensional DCT coefficients Cf1i (i = 1 to 7) are normalized by the following equation (5). The obtained coefficient is the normalized DCT coefficient Cfn1i. In the example of FIG. 6, j = 01.

  For each basic pattern Pj (j is 01 to 07, 11 to 15, 21 to 23, 31), these seven normalized DCT coefficients Cfnji (i = 1 to 7) are stored in the ROM 160 (see FIG. 1) in advance. Has been. These normalized DCT coefficients Cfnji are used in the calculation of the above equation (4).

  FIG. 7 is a diagram showing 16 basic patterns Pj of luminance change (j is 01 to 07, 11 to 15, 21 to 23, 31). In each basic pattern Pj, the vertical axis represents the luminance level, and the horizontal axis represents the pixel position. Each basic pattern Pj represents the luminance level at eight pixel positions, similarly to P01 shown in FIG. For example, the basic pattern P02 is a pattern in which the luminance level at the pixel position 0 is 2, the luminance levels at the pixel positions 2 to 7 are 0, and the luminance level decreases linearly from the pixel position 0 toward 2. . The basic pattern P07 is a pattern in which the luminance level at the pixel position 0 is 7, the luminance level at the pixel position 7 is 0, and the luminance level decreases linearly from the pixel position 0 toward 7.

  The basic patterns P01, P11, P21, and P31 shown in the upper part of FIG. 7 are basic patterns in which the luminance level decreases from left to right in the region for one pixel. The blur width Bw = 1 is associated with these basic patterns. In the basic pattern P01, the position where the luminance level decreases is on the left (pixel positions 0 to 1), and in the basic pattern P31, the position where the luminance level decreases is on the right (pixel positions 3 to 4). In the basic patterns P01, P11, P21, and P31, the position at which the luminance level decreases is more to the left in the order of arrangement.

  Similarly, the other basic patterns in FIG. 7 are assigned the blur width Bw equal to the number of pixels in the region where the luminance level changes (decreases) (all shown below the basic pattern Pj in FIG. 7). If the colors of adjacent areas in the image are different, the brightness of each area is usually different. The blur width Bw represents the degree of blur at the boundary between color areas.

  In FIG. 7, basic patterns having the same assigned blur width Bw are arranged in the horizontal direction. In the basic patterns shown side by side in the horizontal direction, the position at which the luminance level decreases is more to the left as it is shown on the left. Further, these basic patterns Pj do not overlap each other even if they are inverted vertically and horizontally.

  After the sum SDj of the absolute values of the differences between the normalization coefficient Fr0i (i = 1 to 7) of the target region and the normalized DCT coefficient Cfnji of the basic pattern Pj of luminance change is calculated in step S350 of FIG. In step S360, it is determined whether or not there is a basic pattern Pj in which SDj is smaller than a predetermined threshold value Thsd.

  If there is no basic pattern Pj in which SDj is smaller than the predetermined threshold Thsd, the basic pattern to be associated with the target area is determined as “unknown” in step S370, and the luminance change pattern of the target area is “Flat”. Such a target area is a “flat area” (see the processing in steps S60 to S80 in FIG. 3). Then, after the process of step S370, the process of step S400 is performed without performing the processes of steps S380 and S390.

  On the other hand, if there is a basic pattern Pj whose SDj is smaller than the predetermined threshold value Thsd, in step S380, the basic pattern with the smallest SDj is associated with the target area. Further, the blur width Bw (see FIG. 7) of the associated basic pattern is set as the horizontal partial blur level Bbh of the target area.

  If the basic pattern Pmat having the smallest SDj with the normalization coefficient Fr0i (i = 1 to 7) of the target area is a basic pattern Pmat, the luminance change pattern of the target area is the basic pattern Pmat or the basic pattern. It is most similar to the luminance change pattern obtained by inverting Pmat vertically or horizontally. Here, “most similar” means that the pattern is similar to another basic pattern or a pattern obtained by inverting another basic pattern vertically or horizontally.

  Next, in step S390 in FIG. 5, the direction of the luminance change pattern of the target region is determined.

  FIG. 8 is a table showing the relationship between the direction of the luminance change pattern and the sign of the DCT coefficient. FIG. 8 shows an example of the basic pattern P01. When the luminance change pattern of the target area is in the same direction as the basic pattern P01 (P01a in FIG. 8), the sign of the DCT coefficient F01 of the target area is positive (indicated by “+” in the figure). Further, the sign of the DCT coefficient F02 in the target area is the same as the sign of the DCT coefficient F01 in the target area (indicated by “=” in the figure) and is positive.

  As shown in FIG. 8, the same pattern P01a as the basic pattern P01, and patterns P01b to P01d obtained by inverting the basic pattern P01 vertically and horizontally, respectively, have a DCT coefficient F01 sign (+/−) and a DCT coefficient F01. And the difference in the sign of F02 (= / ≠) is different. Although an example of the basic pattern P01 is shown here, the same applies to other basic patterns Pj. Therefore, in step S390 of FIG. 5, the direction of the luminance change pattern can be determined based on the signs of the DCT coefficients F01 and F02 of the target region.

  In step S400 of FIG. 5, it is determined whether or not the processing in steps S310 to S390 has been executed for all target regions. If there is a target area for which the processes of steps S310 to S390 have not yet been performed, the process returns to step S310. In step S310, a target region that has not yet been processed is selected, and its DCT coefficients F00 to F07 are acquired. On the other hand, when the processing of steps S310 to S390 has been executed for all target regions, the edge pattern matching processing is terminated.

  When edge pattern matching processing is performed in the horizontal direction, the processing order of the target area is, in principle, the order from left to right in the image. The leftmost target area is processed. Further, when edge pattern matching processing is performed in the vertical direction, the processing order of the target area is, in principle, the order from the top to the bottom in the image. The target area at the top of is processed.

  FIG. 9 is a flowchart showing the contents of processing executed when the determination result of step S330 of FIG. 5 is No. The process of FIG. 9 is executed when the sum Sfh of the absolute values of the DCT coefficients F01 to F07 of the AC component is equal to or smaller than a predetermined threshold value Thsf (step S330). The case where Sfh is equal to or smaller than the threshold value Thsf means that the variation in luminance in the horizontal direction is within a certain level within the target region.

  In step S510, it is determined whether or not the target area Atf processed immediately before the target area currently processed is a “flat area”.

  If the target area Atf processed immediately before is not a flat area, the luminance change pattern of the target area Atp currently processed is determined to be “flat” in step S520. That is, the target area Atp is assumed to be a “flat area”.

  Note that the process also proceeds from step S510 to S520 when the target region currently being processed is the target region at the beginning (left end) of the row. If the target region currently being processed is the target region at the beginning (left end) of the row, the target region Atf processed immediately before is the target region at the right end of the row one row above, and is currently processed. It is not the target area on the left side of the target area Atp.

  In the case where the process is performed in the vertical direction, the process also proceeds from step S510 to S520 when the target area being processed is the top (top) target area of the column. When the target area currently processed is the target area at the top (upper end) of the column, the target area Atf processed immediately before is the target area at the lower end of the left line, and is currently processed. It is not the target area above the target area Atp.

  On the other hand, when the target region Atf processed immediately before is a “flat region”, in step S530, the DCT of each DC component of the target region Atp currently processed and the target region Atf processed immediately before are obtained. A difference Dav of the coefficient F00 is calculated. Dav represents the difference between the average brightness values of the target area Atp currently processed and the target area Atf processed immediately before. In step S540, it is determined whether or not Dav is greater than a predetermined threshold value Thda.

  When the difference Dav between the average luminance values is equal to or smaller than the predetermined threshold value Thda, in step S520, the luminance change pattern of the target region Atp currently processed is determined to be “flat”. In this case, the image of the target area Atp that is currently processed and the image of the target area Atf that has been processed immediately before are not represented by color boundaries and are represented by luminances that approximate each other. means. A schematic diagram showing such a state is shown on the left side of step S540.

  On the other hand, if Dav is larger than the predetermined threshold value Thda, the basic pattern P01 is associated with the target area Atp currently processed in step S550. Then, the horizontal partial blurring degree Bbh of the target area Atp currently processed is set to 1 which is the blurring width Bw of the basic pattern P01. In this case, it means that the image of the target area Atp currently processed and the image of the target area Atf processed immediately before are represented by luminances that do not approximate each other. That is, it means that the boundary between the target area Atp and the target area Atf is just a color boundary. A schematic diagram showing such a state is shown on the right side of step S540.

  By performing the processing of steps S510 to S550 as described above, even when the edge of the color area in the image is not inside each target area, it is located at the boundary between the adjacent target areas ( 9), the blur width can be determined.

  After step S520 or S550, the process proceeds to step S400 in FIG. 5 (see FIG. 9 and FIG. 5B). The above is the content of the edge pattern matching process.

  The contents of the edge pattern matching processing (see FIGS. 5 to 9) have been described above by taking the case of performing processing in the horizontal direction as an example. However, the same processing can be performed when processing is performed in the vertical direction.

  FIG. 10 is a flowchart showing the contents of edge connection processing in step S30 of FIG. Here, a case where edge connection processing is performed in the horizontal direction will be described as an example. The same processing is performed when edge connection processing is performed in the vertical direction.

  Note that when edge connection processing is performed in the horizontal direction, the processing order of the target area is, in principle, the order from the left to the right in the image. The leftmost target area is processed. In addition, when edge connection processing is performed in the vertical direction, the processing order of the target area is, in principle, the order from the top to the bottom in the image. The upper right target area is processed.

  In step S710 in FIG. 10, the basic pattern associated with a certain target region and the direction of the luminance change pattern are acquired (see steps S380 and S390 in FIG. 5).

  In step S720, it is examined whether or not the basic pattern associated with the target region is P01 to P07 or P11 to P15 (see FIG. 7). The basic patterns P01 to P07 are basic patterns in which the position where the luminance changes is the left end of the pattern. The basic patterns P11 to P15 are basic patterns in which there is a portion where the luminance is constant by one pixel at the left end, and the luminance changes from the pixel position on the right. If the basic pattern associated with the target area is not P01 to P07 or P11 to P15, the process proceeds to step S780.

  If the basic pattern associated with the target area is P01 to P07 or P11 to P15 in step S720, the process proceeds to step S730. In step S730, it is examined whether or not the basic patterns associated with the target region Atf processed immediately before are the basic patterns P01 to P07. If the examination result is No, the process proceeds to step S780. Note that the process also proceeds from step S730 to step S780 when the target area currently being processed is the target area at the beginning (left end) of the line.

  If the basic pattern associated with the target area Atf processed immediately before is P01 to P07, the process proceeds to step S740. In step S740, it is examined whether or not the position where the luminance changes is opposite between the currently processed target area Atp and the immediately previous processed target area Atf. Note that the target region Atf processed immediately before is a target region adjacent to the left side of the target region Atp currently processed. If the examination result is No, the process proceeds to step S780.

  For example, when the position where the luminance changes in both the target area Atf and the target area Atp is at the left end, the determination result in step S740 is No. A schematic diagram showing such a case is shown on the left side of step S740 in FIG. 10 using the luminance change pattern Pf of the target region Atf and the luminance change pattern Pp of the target region Atp. On the other hand, a schematic diagram in the case where the positions where the luminance changes in the target area Atf and the target area Atp face each other is shown on the lower right side of step S740.

  In step S740, if the target region Atp currently processed and the target region Atf processed immediately before are located at positions where the luminance changes, the process proceeds to step S750. In step S750, it is examined whether the direction of the luminance change of the target area Atp currently processed matches the direction of the luminance change of the target area Atf processed immediately before. “Luminance change directions are the same” means that the luminance decreases from left to right, or the luminance increases from left to right. If the examination result is No, the process proceeds to step S760.

  The case where the examination result in step S750 is No is a case where the luminance changes to the V-type or the Λ-type across the boundary of the symmetric region. For example, when the luminance decreases from the left to the right in the target region Atf and the luminance increases from the left to the right in the target region Atp, the luminance change is V-type, and the determination result in step S750 is No. Become. A schematic diagram showing such a case is shown on the left side of step S750 in the figure. On the other hand, a schematic diagram in the case where the direction in which the luminance changes in the target area Atf and the target area Atp is the same is shown on the lower right side of step S740 in the figure.

  In step S760, the luminance change pattern of the target region is replaced with “flat”. That is, the target area is a “flat area”. Then, the horizontal blur level Bbh of the target area is cleared (see step S380 in FIG. 5). By performing such processing, it is possible to classify a target region whose luminance is finely increased and decreased as a flat region. As a result, it is possible to exclude a target area whose luminance is finely increased or decreased when determining the focus of the image (see the processing in steps S60 to S80 in FIG. 3).

  When human hair or animal hair is shown in the photograph, the brightness of the portion changes finely. Therefore, the partial blurring degree Bb is small in the target region in which human or animal hair is reflected (see step S50 in FIG. 3). As a result, when human or animal hair appears in a certain area or more in the photographic image, it is always determined that the image is a focused image in the processing of steps S70 and S80 in FIG. there is a possibility. However, in the present embodiment, by performing the processing in steps S750 and S760, such a target area is excluded from the focus determination target as a flat area. For this reason, it is possible to accurately determine the focus of the image.

  In step S750 of FIG. 10, when the direction of the luminance change of the target area Atp currently processed matches the direction of the luminance change of the target area Atf processed immediately before, the process proceeds to step S770. Proceed to That is, the process of step S770 is selected based on the section where the luminance changes in the first selected basic pattern data selected based on the target area to be processed and the target area processed immediately before. In the second selected basic pattern data, the section in which the luminance changes is located at the end of the partner target area, and the direction of the luminance change is the same.

  In step S770, the horizontal partial blur level Bbh of the target area Atp currently processed and the target area Atf processed immediately before is replaced with a new value. Specifically, the sum of the blur width Bw of the basic pattern associated with the target area Atp and the blur width Bw of the basic pattern associated with the target area Atf is newly added to the target area Atp and the target. The horizontal partial blurring degree Bbh of the region Atf is assumed.

  In step S780, it is determined whether or not the processing in steps S710 to S770 has been executed for all target regions. If there is a target area for which the processes of steps S710 to S770 have not been performed, the process returns to step S710. In step S710, for a target area that has not yet been processed, the basic pattern associated with the target area and the direction of the luminance change pattern are acquired. On the other hand, when the processes of steps S710 to S770 have been executed for all target areas, the edge linking process is terminated.

  By performing the edge linking process as described above, the blur width can be correctly determined even when the edge (contour) of the color area in the image exists across a plurality of target areas.

  Also, when calculating the luminance change pattern of the target area in the image, even if the luminance continues to change across the target area, the edge of the area (target An area having a constant luminance for one pixel may occur at the boundary between the areas. However, in the edge linking process according to the present embodiment, when there is a constant luminance region for one pixel at the end of the target region, the processing is performed in the same manner as when the luminance continuously changes (FIG. 10). Step S720). Specifically, there is a portion where the luminance is constant by one pixel at the left end, and the basic patterns P11 to P15 whose luminance changes from the pixel position on the right are associated with the target region currently being processed. Are also subjected to edge connection processing in step S770. For this reason, it is possible to correctly determine the blur width of the edge (outline) of the color region in the image.

  In the above description, the contents of the edge linking process have been described by taking the case of performing processing in the horizontal direction as an example. However, the same processing can be performed when processing is performed in the vertical direction.

  In the case of processing in the vertical direction, when handling the data of the target region that has been processed immediately before (see step S730 in FIG. 10), the processing is performed as follows. That is, in the edge linking process, when the target area being processed is the top (top) target area of the column, the process proceeds from step S730 to S780. When the target area currently processed is the target area at the top (upper end) of the column, the target area Atf processed immediately before is the target area at the lower end of the left line, and is currently processed. It is not the target area above the target area Atp.

B. Second embodiment:
In the second embodiment, the basic pattern (see FIG. 7) stored in the ROM 160 is different from the first embodiment. The second embodiment is different from the first embodiment in part of edge pattern matching processing (see S20 in FIG. 3 and FIG. 5). Further, in the second embodiment, when the luminance difference at the color boundary portion is equal to or smaller than a predetermined value, it is assumed that there is no luminance difference. The other points of the second embodiment are the same as those of the first embodiment. Hereinafter, the second embodiment will be described focusing on the differences from the first embodiment.

  FIGS. 11 to 14 are diagrams showing basic pattern groups PGa to PGd each including 16 basic patterns of luminance change. The notation of FIGS. 11-14 is the same as FIG. In FIGS. 11 to 14, basic patterns of Bw = 3 to 6 (corresponding to j = 03 to 06, 13 to 15, 23) are omitted.

  A basic pattern Pja (j represents 01 to 07, 11 to 15, 21 to 23, 31) of the basic pattern group PGa shown in FIG. 11 is a basic pattern in which the difference between the maximum luminance and the minimum luminance is equal to each other. . The magnitude of this luminance difference is defined as a unit luminance difference UL. The unit luminance difference UL can be set to an amount corresponding to luminance 1, for example. Other points of the basic pattern Pja of the basic pattern group PGa are the same as the basic pattern Pj (see FIG. 7) of the first embodiment. The basic pattern group PGa is a group of basic patterns in which there is a luminance change that decreases toward the right at the first pixel position on the left side, and the luminance is constant at the pixel position on the right side of the first pixel position.

  For example, in the first embodiment, as shown in FIG. 6, the luminance level L01i (i = 0 to 7) at each pixel position i (i = 0 to 7) of the basic pattern P01 is 1, 0, 0 in order. , 0, 0, 0, 0. On the other hand, the luminance levels L01ai (i = 0 to 7) at the pixel positions i (i = 0 to 7) of the basic pattern P01a shown in the upper left in FIG. 11 are UL, 0, 0, 0, 0, 0, 0 (see FIG. 6). In FIG. 11, the luminance level L02ai (i = 0-7) at each pixel position i (i = 0-7) of the basic pattern P02a shown below the basic pattern P01a is UL, UL / 2, 0 in this order. , 0, 0, 0, 0.

  In the ROM 160 of the printer 1, signed normalized DCT coefficients Cfsjai (j = 01 to 07, 11 to 15, 21 to 23, 31, i = 0 to 7) of the basic pattern Pja are stored. Similar to the normalized DCT coefficient Cfnji of the first embodiment, these signed normalized DCT coefficients Cfsjai are Ljai (j = 01 to 07, 11 to 15, 21 to 23, 31, i = 0) of the basic pattern Pja. ~ 7).

  That is, a one-dimensional DCT coefficient Cfjai (i = 0 to 7) is obtained by performing a one-dimensional DCT on the luminance level Ljai (i = 0 to 7). Of the eight one-dimensional DCT coefficients Cfjai, the absolute value of the seven one-dimensional DCT coefficients Cfjai is “signed normalized” by the following equation (6), so that the signed normalized DCT coefficient Cfsjai is Is obtained (see FIG. 6). The numerator on the right side of Equation (6) is not given an absolute value symbol.

  In this way, for each of a plurality of coefficients, each coefficient is divided by the sum of the absolute values of the coefficients, and “a plurality of new coefficients having the same number as the original coefficient and the sum of the absolute values being 1” is obtained. The calculation of the “coefficient” is referred to herein as “signed normalization”.

  In addition, the ROM 160 of the printer 1 stores a sum Sja of absolute values of one-dimensional DCT coefficients Cfjai (i = 1 to 7) that are AC components among the one-dimensional DCT coefficients Cfjai of the basic pattern Pja. Sja is calculated by the following equation (7).

  The basic patterns Pjb (j = 01 to 07, 11 to 15, 21 to 23, 31) of the basic pattern group PGb shown in FIG. 12 are the basic patterns Pja (j = 01 to 07, 11 to 15, 21 to 23, 31) are patterns that are reversed left and right. For reference, the basic pattern Pja of the basic pattern group PGa is indicated by a broken line in FIG. The basic pattern group PGb is a group of basic patterns in which there is a luminance change that increases toward the right at the first pixel position on the right side, and the luminance is constant at the pixel position on the left side of the first pixel position.

  The signed normalized DCT coefficients Cfsjbi (i = 1 to 7) of the basic pattern Pjb of the basic pattern group PGb are respectively the signed normalized DCT coefficients Cfsjai (i = 1 to 7) of the basic pattern Pja of the basic pattern group PGa. Of these, it is obtained by reversing the sign of the odd component.

  The basic patterns Pjc (j = 01 to 07, 11 to 15, 21 to 23, 31) of the basic pattern group PGc shown in FIG. 13 are the basic patterns Pja (j = 01 to 07, 11 to 15, 21 to 23, 31). For reference, the basic pattern Pja of the basic pattern group PGa is indicated by a broken line in FIG. The basic pattern group PGc is a group of basic patterns in which there is a luminance change that increases toward the right at the first pixel position on the left side, and the luminance is constant at the pixel position on the right side of the first pixel position.

  The signed normalized DCT coefficients Cfsjci (i = 1 to 7) of the basic pattern Pjc of the basic pattern group PGc are respectively the signed normalized DCT coefficients Cfsjai (i = 0 to 7) of the basic pattern Pja of the basic pattern group PGa. It is obtained by reversing the sign of positive and negative.

  The basic patterns Pjd (j = 01 to 07, 11 to 15, 21 to 23, 31) of the basic pattern group PGd shown in FIG. 14 are the basic patterns Pja (j = 01 to 07, 11 to 15, 21 to 23, 31) are patterns that are inverted horizontally and vertically. For reference, the basic pattern Pja of the basic pattern group PGa is indicated by a broken line in FIG. The basic pattern group PGd is a group of basic patterns in which there is a luminance change that decreases toward the right at the first pixel position on the right side, and the luminance is constant at the pixel position on the left side of the first pixel position.

  The signed normalized DCT coefficients Cfsjdi (i = 1 to 7) of the basic pattern Pjd of the basic pattern group PGd are respectively the signed normalized DCT coefficients Cfsjai (i = 1 to 7) of the basic pattern Pja of the basic pattern group PGa. Of these, it is obtained by reversing the signs of odd-numbered components and reversing the signs of all the coefficients. That is, the signed normalized DCT coefficient Cfsjdi (i = 1 to 7) of the basic pattern Pjd of the basic pattern group PGd is the signed normalized DCT coefficient Cfsjai (i = 1 to 7) of the basic pattern Pja of the basic pattern group PGa. ) Of the even-numbered components.

  In the ROM 160 of the printer 1, for the basic patterns Pjb, Pjc, and Pjd of these basic pattern groups PGb, PGc, and PGd, signed normalized DCT coefficients Cfsjbi, Cfsjci, Cfsjdi, and the sum of absolute values of AC components Sjb, Sjc and Sjd are stored (j = 01-07, 11-15, 21-23, 31, i = 1-7). The calculation method of the signed normalized DCT coefficients Cfsjbi, Cfsjci, Cfsjdi and the sum of absolute values of AC components Sjb, Sjc, Sjd is the same as that of the basic pattern group PGa.

  15 and 16 are flowcharts showing the processing contents of edge pattern matching in the second embodiment. 15 and 16, steps for performing the same processing as in FIG. 5 of the first embodiment are displayed using the same reference numerals as in FIG. 5. Further, the illustration and description of the steps for performing the same processing as in FIG. 5 of the first embodiment may be omitted.

  In step S310 of FIG. 15, eight DCT coefficients F00 to F07 in the horizontal direction of a certain target region are acquired. The processing in step S310 is the same as that in the first embodiment.

  In step S315, based on the DCT coefficients F01 and F02 which are AC components of the first and second lowest frequencies among the acquired eight DCT coefficients F00 to F07, one of the basic pattern groups PGa to PGd is selected. One basic pattern group is selected.

  As described above, the direction of the luminance change pattern can be specified based on the DCT coefficients F01 and F02 (see FIG. 8). For this reason, in step S315, based on the DCT coefficients F01 and F02, a basic pattern group is selected according to the direction of the luminance change pattern. The processing in step S315 is performed by a group selection unit as a functional unit of the CPU. The group selection unit is included in the basic pattern selection unit 120 (see FIG. 1) as a function unit of the CPU.

  FIG. 17 is a table showing the basic pattern group selection method in step S315. The notation in the left two columns in FIG. 17 is the same as the notation in the right two columns in FIG. In step S315, a basic pattern group is selected as shown in FIG. 17 depending on whether the sign of DCT coefficient F01 and the signs of DCT coefficients F01 and F02 are equal.

  For example, as shown in the uppermost stage of FIG. 17, when the sign of the DCT coefficient F01 in the target region is positive and the sign of the DCT coefficient F02 is positive as well as the sign of the DCT coefficient F01, the luminance change pattern is There is a luminance change that decreases toward the right at the pixel position, and it is considered that the pixel position on the right side is close to a shape with constant luminance (see FIG. 8). For this reason, there is a luminance change that decreases toward the right at the left pixel position, and a basic pattern group PGa having a basic pattern with a constant luminance at the right pixel position is selected (see FIG. 11).

  By performing the process of step S315, it is possible to perform pattern matching by selecting a basic pattern group having similar brightness change pattern trends in advance. That is, it is possible to limit the number of basic patterns to be subjected to pattern matching in advance. Therefore, the processing load can be reduced as compared with the aspect in which the coefficients are compared for the basic patterns Pja to Pjd of all the basic pattern groups PGa to PGd.

  In step S320 of FIG. 15, the sum Sfh of absolute values of the DCT coefficients F01 to F07 of the AC component of the target region is calculated. The process of step S320 of the second embodiment and the subsequent step S330 are the same as those of the first embodiment.

  In step S345, signed normalized coefficients Fs01 to Fs07 are calculated according to the following equation (8) based on the DCT coefficients F01 to F07 of the AC components in the horizontal direction of the target region. Expression (8) is an expression for calculating signed normalized coefficients Fs01 to Fs07 corresponding to the DCT coefficients F01 to F07, respectively, so that the sum of absolute values becomes 1 based on the DCT coefficients F01 to F07. In addition, i is an integer of 1-7.

  In step S355 of FIG. 15, SDsj is calculated according to the following equation (9). That is, the signed normalization coefficient Fs0i (i = 1 to 7) of the target region and the basic pattern of the luminance change of the basic pattern group (PGx, where x is a, b, c, or d) selected in step S315. The difference between the normalized DCT coefficients Cfsjxi (x is a, b, c or d) of Pjx (j is 01 to 07, 11 to 15, 21 to 23, 31. x is a, b, c or d) is calculated. Is done. Then, the sum SDsj of the absolute values of the differences is calculated (j is 01 to 07, 11 to 15, 21 to 23, 31).

  In step S365, it is determined whether or not there is a basic pattern Pj in which SDsj is smaller than a predetermined threshold value Thsd.

  If there is no basic pattern Pj in which SDsj is smaller than the predetermined threshold Thsd, the process proceeds to step S370 (see FIG. 5). That is, the basic pattern to be associated with the target area is “unknown”, and the luminance change pattern of the target area is “flat”. The process in step S370 is the same as that in the first embodiment. Thereafter, the processing proceeds to step S400 in step FIG.

  On the other hand, if there is a basic pattern Pj with SDsj smaller than the predetermined threshold Thsd, the basic pattern with the smallest SDsj is associated with the target area in step S381 of FIG. Further, the blur width Bw (see FIGS. 11 to 14) of the associated basic pattern is set as the horizontal partial blur level Bbh of the target region. The process of step S381 is the same as the process of step S380 in FIG. 5 except that the determination is based on SDsj (see formula (9)) instead of SDj (see formula (4)).

  In the process of the first embodiment, pattern matching is performed using the absolute value of the DCT coefficient F0i (i = 1 to 7) of the AC component of the target region (see Expression (3)). For this reason, there is a possibility that the same basic pattern may be associated with a plurality of different brightness change patterns. For example, according to equation (3), Fr02 has the same value regardless of whether F02 is positive or negative. Therefore, there is a possibility that the luminance change pattern in which F01, F02, and F03 are all positive and the luminance change pattern in which F01 and F03 are positive but F02 are negative are treated as the same pattern. is there.

  In the second embodiment, four basic pattern groups having different luminance change patterns are prepared (see FIGS. 11 to 14), and the basic pattern groups to be subjected to pattern matching based on the luminance change patterns of the target region. Is selected (see step S315 in FIG. 15). Then, signed normalization is performed so as not to lose the information of the positive and negative signs (see Expression (8)), and pattern matching is performed based on the signed normalization coefficient Fs0i (i = 1 to 7) (FIG. 15). Step S345-see S381 in FIG. For this reason, although it is actually a different brightness | luminance change pattern, possibility that the same basic pattern will be matched can be reduced.

  In step S383 of FIG. 16, an evaluation value Ld corresponding to the difference between the maximum value and the minimum value of the luminance of the target region is calculated and stored in the main memory 150. The evaluation value Ld is obtained by the following formula (10). The evaluation value Ld is an approximate value of the difference between the maximum value and the minimum value of the luminance of the target area.

  Ld = {Sfh / (√2) Smat} × UL (10)

  Smat is the sum of the absolute values of the AC components of the basic pattern Pjx (x is a, b, c, or d) associated in step S381. As already described, the sum Sja, Sjb, Sjc, Sjd of the absolute values of the AC components of the basic patterns Pja, Pjb, Pjc, Pjd is stored in the ROM 160 in advance. UL is the difference between the maximum luminance and the minimum luminance of the basic patterns Pja, Pjb, Pjc, Pjd (see FIGS. 11 to 14).

  The shape of the luminance change pattern in the target region is similar to the basic pattern Pjx associated in step S381. The luminance change pattern of the target area is considered to be close to the luminance change obtained by multiplying the basic pattern Pjx associated in step S381 by a constant with respect to the luminance. For this reason, “the sum Sfh of the absolute values of the DCT coefficients F01 to F07 of the AC component of the target region” is changed to “the one-dimensional DCT coefficient Cfjxi (i = 1 to 7) of the AC component of the basic pattern Pjx associated in step S381”. ) Is multiplied by the luminance difference UL of the basic pattern Pjx to obtain an approximate value Ld of the difference between the maximum luminance and the minimum luminance of the target region. can get. Hereinafter, the evaluation value Ld may be expressed as “luminance difference Ld”.

  The DCT coefficient held by the image data in the JPEG format is a DCT coefficient obtained by two-dimensional DCT (see FIG. 4). On the other hand, the DCT coefficient Cfjxi (j is 01 to 07, 11 to 15, 21 to 23, 31, x is a, b, c, or d, i is 1 to 7) of each basic pattern Pja is a luminance. This is a DCT coefficient obtained by performing a one-dimensional DCT on the pattern of change. Therefore, by dividing “the sum Sfh of the absolute values of the DCT coefficients F01 to F07” by the “sum of the absolute values of the one-dimensional DCT coefficients Cfjxi (i = 1 to 7) of the AC components of the basic pattern Pjx” itself. Thus, the approximate luminance difference Ld cannot be obtained.

  Based on the luminance L (m, n) (m = 0 to 7, n = 0 to 7) of each pixel in the target region, a two-dimensional DCT coefficient Fuv (u = 0 to 7, v = 0 to 7) is obtained. The calculation formula (11) at that time is as follows.

  Here, when calculating the values of F01 to F07, that is, when u = 0, Expression (11) is as follows. Note that aL (n) represents the average luminance in the vertical (column) direction of the (n + 1) -th column in the 8 × 8 target region.

  On the other hand, the one-dimensional DCT coefficient Cfjxi is calculated based on the luminance Ljxi (i = 0 to 7, x is a, b, c, or d. J can be 01 to 07, 11 to 15, 21 to 23, 31). The calculation formula (13) for obtaining is as follows.

  As can be seen from the comparison between Expression (12) and Expression (13), F0v obtained from Expression (12) is a coefficient value obtained by performing one-dimensional DCT on the average luminance aL (n) in the column direction. It is understood that it is equal to 2√2 times.

  Therefore, “the sum Sfh of the absolute values of the DCT coefficients F01 to F07” is obtained by multiplying the sum Smat of the absolute values of the one-dimensional DCT coefficients Cfjxi (i = 1 to 7) of the AC component of the basic pattern Pjx by 2√2. Is multiplied by the luminance difference UL of the basic pattern Pjx to obtain an approximate value Ld of the difference between the maximum luminance and the minimum luminance of the target region. Note that the relationship between the two-dimensional DCT coefficient and the one-dimensional DCT coefficient described above also holds for the two-dimensional DCT coefficient Fuv with u = 1 to 7 and v = 0.

  When the luminance difference Ld of the target area is obtained in step S383 as described above, after that, the process of step S400 is performed without performing the process of S390 (see FIG. 5). The processing in step S400 is the same as in the first embodiment.

  Note that the information of “the direction of the luminance change pattern of the target region” obtained in step S390 in the first example is obtained in step S315 (FIG. 15) in the second example. In the second embodiment, when the process is performed using the information “the direction of the luminance change pattern of the target area”, the process is performed based on the information obtained and stored in step S315. .

  18 and 19 are flowcharts showing the contents of edge connection processing in the second embodiment. The flowcharts of FIGS. 18 and 19 differ in the basic pattern considered in the process of step S721 corresponding to step S720 in the first embodiment and the process of step S731 corresponding to step S730 in the first embodiment. Further, the flowcharts of FIGS. 18 and 19 have the determination and branching processing of step S755. The other points of the flowcharts of FIGS. 18 and 19 are the same as the flowchart of FIG. 10 of the first embodiment. Also in the second embodiment, a case where edge connection processing is performed in the horizontal direction will be described as an example. The same processing is performed when edge connection processing is performed in the vertical direction.

  Step S721 in FIG. 18 is processing corresponding to step S720 in FIG. In step S721, it is determined whether or not the basic pattern associated with the target region is a pattern corresponding to P01 to P07 or P11 to P15 in the first example in the basic pattern groups PGa to PGd. More specifically, the basic patterns associated with the target area are P01a to P07a or P11a to P15a, P01b to P07b or P11b to P15b, P01c to P07c or P11c to P15c, or P01d to P07d or P11d to P15d. It is determined whether or not there is. If the basic patterns associated with the target area are those patterns, the process proceeds to step S731. Otherwise, the process proceeds to step S780.

  Step S731 is a process corresponding to step S730 in FIG. In step S731, whether or not the basic pattern associated with the target area Atf processed immediately before is a pattern corresponding to the basic patterns P01 to P07 in the first embodiment in the basic pattern groups PGa to PGd. Be considered. More specifically, it is determined whether the basic pattern associated with the target area Atf processed immediately before is the basic pattern P01a to P07a, P01b to P07b, P01c to P07c, or P01d to P07d. . If the basic patterns associated with the immediately preceding target area Atf are those patterns, the process proceeds to step S740. Otherwise, the process proceeds to step S780. Note that the process also proceeds to S780 when the target region currently being processed is the target region at the beginning of the line (the left end of the image).

  The processing in step S740 in FIG. 18 and step S750 in FIG. 19 is the same as that in the first embodiment (see FIG. 10), and thus the description thereof is omitted.

  In step S755, the sum of the luminance difference Ld of the target region Atp (see step S383 in FIG. 16) and the cumulative luminance difference SLdf of the target region Atf processed immediately before is calculated as the cumulative luminance difference SLd of the target region Atp. . If the target area Atp is the leftmost target area in the image, the luminance difference Ld of the target area Atp is set as the cumulative luminance difference SLd. The cumulative luminance difference SLd of the target area Atp is the luminance for a plurality of target areas that exist continuously in the horizontal direction, and the two target areas that are adjacent to each other satisfy the conditions of steps S721 to S750. This is the total value of the differences Ld.

  In step 755, it is determined whether or not the accumulated luminance difference SLd is larger than a predetermined threshold value Thld. If the accumulated luminance difference SLd is greater than the threshold value Thld, the process proceeds to step S770. Also, when the target area currently processed is the target area at the head (left end) of the line, the process proceeds from step S755 to S770. The processing after step S770 is the same as that in the first embodiment (see FIG. 10).

  On the other hand, when the accumulated luminance difference SLd is equal to or smaller than the threshold value Thld, the process proceeds from step S755 to step S760. In step S760, the luminance change pattern of the target region is replaced with “flat”. That is, the target area is a “flat area”. Then, the horizontal blur level Bbh of the target area is cleared (see step S381 in FIG. 16). Similarly, the accumulated luminance difference SLd of the target area is also cleared. The subsequent processing is the same as in the first embodiment (see FIG. 5). Note that the case where SLd is equal to or less than the threshold value Thld means that the difference in luminance at the connected edges (between different color regions) is less than or equal to some extent.

  In the second embodiment, by performing such processing, it is possible to classify a target area whose edge luminance difference in the horizontal direction is a certain level or less into a flat area. As a result, a portion where the color or brightness changes gently in the image and is not a boundary (that is, a contour) between regions having different colors is treated as a contour to determine the focus (FIG. 3). (See the processing in steps S60 to S80). As a result, the accuracy of in-focus determination can be increased.

  In the present embodiment, the luminance difference Ld of the target area is calculated based on the DCT coefficients F01 to F07 of the AC component of the target area and the AC component of the basic pattern associated with the target area ( (See step S383). For this reason, compared with the aspect which performs inverse DCT (Inverse Discrete Cosine Transform) transformation based on the DCT coefficients F00 to F07 of the target region, calculates the luminance of each pixel of the target region, and then calculates the luminance difference, The processing load is small. As a result, the time required for processing can be shortened.

  In the above description, the content of the edge pattern matching process has been described by taking the case of performing the process in the horizontal direction as an example. However, the same processing can be performed when processing is performed in the vertical direction.

C. Variations:
The present invention is not limited to the above-described examples and embodiments, and can be implemented in various modes without departing from the gist thereof. For example, the following modifications are possible.

C1. Modification 1:
In the above embodiment, the edge pattern matching process is performed based on the luminance pattern of each pixel. However, the edge pattern matching process can also be performed based on other gradation values of each pixel. For example, instead of luminance, edge pattern matching is performed based on either the intensity of each color component in the RGB color system, or the gradation value representing the intensity of Cr or Cb in the YCrCb color system. Processing can also be performed. That is, the edge pattern matching process can be performed based on data representing a color regardless of whether it is a chromatic color having saturation and hue and an achromatic color that can be represented only by luminance (lightness). it can.

C2. Modification 2:
When performing the edge pattern matching process, the following process can also be performed. That is, all the basic patterns P01 to P31 are prioritized. This order is, for example, the order in which the basic patterns are ranked in order from the one that is most likely to be selected in step S380 in FIG. 5 using a predetermined number of photographic images as samples. This ranking can also be performed according to the scene of the photograph, such as “person”, “landscape”, “night view”, and the like.

  Then, when performing the edge pattern matching processing, the absolute value of the difference between the normalized DCT coefficient Cfnji of the basic pattern and the normalized coefficient Fr0i (i = 1 to 7) of the target region in order according to the order. (See step S350 and equation (4) in FIG. 5). Then, the first basic pattern having SDj smaller than a predetermined reference is associated with the target area. Note that SDj is not calculated for a basic pattern having a lower rank than the associated basic pattern.

  With such an aspect, it is possible to reduce the processing load while suppressing a decrease in the accuracy of the edge pattern matching processing. Although the description has been given based on the first embodiment, the same processing can be executed by ranking the basic patterns in each basic pattern group in the second embodiment.

C3. Modification 3:
In the first embodiment, SDj that is “the sum of absolute values of differences between the normalization coefficient Fr0i (i = 1 to 7) of the target region and the normalized DCT coefficient Cfnji of the luminance change basic pattern Pj” is the smallest. The basic pattern is associated with the target area. However, the basic pattern can be selected in other ways.

  For example, the basic pattern having the smallest “average of absolute values of differences between the normalization coefficient of the target area and the normalized DCT coefficient Cfnji of the basic pattern Pj” may be associated with the target area. Further, the basic pattern having the smallest “sum of squares of differences between the normalization coefficient of the target area and the normalized DCT coefficient Cfnji of the basic pattern Pj” may be associated with the target area. Furthermore, the basic pattern having the smallest “average squared difference between the normalization coefficient of the target region and the normalized DCT coefficient Cfnji of the basic pattern Pj” may be associated with the target region.

  That is, based on the difference between the coefficient representing the color change pattern of the target area and the basic coefficient of the basic pattern, the basic pattern having the smallest difference between the two basic patterns or the difference between the two is predetermined. A basic pattern smaller than the standard can be selected as a basic pattern associated with the target area. Although the description has been given based on the first embodiment, the basic pattern can be similarly selected for the second embodiment by various methods.

C4. Modification 4:
In the first embodiment, each basic pattern holds a normalized DCT coefficient (see FIG. 6), and the DCT coefficient of the image data is normalized and then compared with the normalized DCT coefficient of each basic pattern. (See steps S340 and S350 in FIG. 5). However, the basic coefficient corresponding to each frequency component of each basic pattern can be held without normalization. The DCT coefficients of the image data can be compared with those basic coefficients without normalization.

  Similarly, also in the second embodiment, the basic coefficient corresponding to each frequency component of each basic pattern can be held without being signed and normalized. Then, the DCT coefficients of the image data can be compared with those basic coefficients without being normalized with a sign.

C5. Modification 5:
In the first embodiment, in the edge linking process, it is examined whether or not the basic pattern associated with the target region processed immediately before is a specific basic pattern P01 to P07 (step in FIG. 10). S730). However, in step S730, it is also possible to consider whether the basic pattern associated with the target region processed immediately before is P01 to P07 or P11 to P15.

  In such an aspect, even when there is a constant luminance region for one pixel at the end of the target region processed immediately before, processing is performed in the same manner as when the luminance continuously changes at the boundary of the target region. Is done. According to such an aspect, even when a region with a constant luminance of one pixel is generated at the end of the region in the immediately preceding target region due to an error in image processing, the end of the color region in the image The blur width of (contour) can be correctly determined.

C6. Modification 6:
In the first embodiment and the modified example 5, not only when the luminance continues to increase or decrease at the boundary of the target region, but also when there is a constant luminance region for one pixel at the end of the target region, The (vertical) partial blur level is modified based on the sum of the partial blur levels of both target regions (see step S720 in FIG. 10). However, even when there is an area with a constant luminance of two pixels or more at the end of the target area, the horizontal (vertical) partial blurring degree can be modified based on the sum of the partial blurring degree of the target area. . The size of the “constant luminance region” serving as a reference at that time can be determined based on the size of the basic pattern and the size (pixel density) when printing the image of the image data.

C7. Modification 7:
In the above embodiment, the basic pattern represents a change in luminance for a section having a width of 8 pixels (see FIGS. 6 and 7). However, it is also possible to prepare a basic pattern that represents a change in luminance for sections of different widths and perform edge pattern matching processing. In such an aspect, after the edge pattern matching process, the edge connection process may not be performed, or the edge connection process may be performed.

C8. Modification 8:
Edge connection processing can be performed for two or more target regions. As modes other than the above-mentioned example, the following modes are also possible, for example. That is, first, by selecting a target area associated with the basic pattern P07 whose luminance changes in all sections, and then repeating the same process as in FIG. 10 for the target areas adjacent to both sides of the target area. Edge connection processing for three or more target regions can be realized.

  Further, when calculating the total value SLd of the luminance differences Ld for a plurality of target regions, comparing SLd with a predetermined threshold value Thld, and making a determination based on the luminance difference (see step S755 in FIG. 18). The following embodiments are preferable. That is, when the process of step S770 in FIG. 18 is performed, the cumulative luminance difference SLd of each target region is stored as the total value of the luminance differences Ld of the target regions. Note that the number of target areas is set to 1 before the process of step S770 is performed for the first time. After that, when performing the process of step S755 for a certain target region, the threshold Thld to be compared is selected according to the number of target regions considered in determining the accumulated luminance difference SLd of the target region. It is preferable to do.

  Moreover, it can also be set as the following aspects. That is, in the processing of FIG. 18 and FIG. 19, in step S755, the cumulative luminance difference SLd is calculated, and the process proceeds to step S770 without performing the determination processing as it is. In this way, the partial blurring degree of each target area is once determined. Then, after each step S780, it is determined whether each partial blur level replaced in step S770 is larger than a predetermined threshold value. When the replaced partial blurring degree is smaller than the predetermined threshold value, the partial blurring degree is cleared. That is, the partial blurring degree is set to zero. Then, the process ends.

  Even in such an aspect, it is possible to reduce a possibility that a portion which is not a boundary between regions having different colors is treated as an outline and considered in the determination of in-focus.

C9. Modification 9:
In the first embodiment, the horizontal (or vertical) partial blur of the current target area Atp and the last processed target area Atf are replaced with new values (see step S770 in FIG. 10). However, the target area where the horizontal (or vertical) partial blurring degree is replaced with a new value is not limited to these target areas. For example, a mode in which the horizontal (or vertical) partial blurring degree of any one of the target area Atp currently processed and the target area Atf processed immediately before can be replaced with a new value.

  Also, when performing edge connection processing for three or more target regions, only one of the three or more target regions subjected to edge connection processing (for example, the first target region or the last target region). The horizontal (or vertical) partial blurring degree may be replaced with a new value. On the other hand, the horizontal (or vertical) partial blurring degree may be replaced with a new value for some or all of the three or more target regions subjected to the edge linking process.

C10. Modification 10:
In the first embodiment, the basic pattern is associated with the flat area (FIG. 9). Further, in the above embodiment, in the edge connection process, when the luminance increases or decreases, the luminance change pattern is replaced with “flat” (step S760 in FIG. 10). However, it is also possible to adopt a mode in which either or both of these processes are not performed. With such an aspect, the processing load can be reduced.

C11. Modification 11:
In the above embodiment, the larger one of the horizontal partial blur level Bbh and the vertical partial blur level Bbv of the target area is set as the partial blur level Bb of the target area (see step S50 in FIG. 3). However, the partial blurring degree Bb of the target region can be determined by other methods.

  For example, the smaller of the horizontal partial blur level Bbh and the vertical partial blur level Bbv can be set as the partial blur level Bb of the target region. Further, an arithmetic average or a geometric average of the horizontal partial blurring degree Bbh and the vertical partial blurring degree Bbv can be used as the partial blurring degree Bb of the target region. At that time, the horizontal partial blurring degree Bbh and the vertical partial blurring degree Bbv can also be weighted. That is, the partial blur level of the target area can be calculated using the horizontal partial blur and the vertical partial blur level of the target area.

C12. Modification 12:
In the above embodiment, each threshold value (for example, threshold value Thbb in step S60 in FIG. 3 or threshold value Thrf in step S80, threshold value Thsf in step S330 in FIG. 5, threshold value Thsd in step S360, The threshold value Thsda in step S540 in FIG. 9) was a fixed value. However, each of these threshold values can also be selected according to the characteristics of the image data.

  For example, it is preferable to select a threshold value used for determination in each step according to the brightness of the image of the image data. This is because even if the difference in luminance (color) is similar, whether or not it is likely to be noticed by humans depends on the absolute size of the luminance. Similarly, even if the width of the boundary when changing from one color to another is about the same, whether or not it is easily noticeable by humans depends on the luminance of the portion where the boundary exists.

  It is also preferable to determine the threshold values according to the color components of red, green, blue, Cr, and Cb. This is because, in different colors, even if they have the same gradation difference and the same edge width, whether or not they are easily noticeable by humans is different.

C13. Modification 13:
In the above embodiment, the ratio Rf obtained by dividing the number Nf of in-focus areas in the image by the sum of the number Nf of in-focus areas and the number Nn of out-of-focus areas is equal to or greater than a predetermined threshold Thrf. Then, the image of the image data file is determined to be a focused image (see steps S70 and S80 in FIG. 3). The in-focus area is an area where the partial blur level Bb is smaller than a predetermined threshold Thbb (see step S60 in FIG. 3). However, the determination of whether or not the image is an in-focus image can be performed by other methods.

  For example, a rectangular area having a predetermined size and having a focal area whose partial blurring degree is smaller than the first reference value is larger than the second reference value in the rectangular area. When present in a photographic image, it can also be determined that the photographic image is a focused image. With such an aspect, an image that is in focus for a part that is particularly important to the user (for example, a face region in which a face is reflected), and an in-focus image even if the other part is not in focus It can be determined that

  Note that the rectangular area can be a rectangular area having a size of Hcm × Wcm when an image is printed on an L plate (8.9 cm × 12.7 cm). H and W can be 0.5 to 1.5, respectively, and preferably 0.8 to 1.2. This is because an object that the user intends to photograph usually has a size larger than that in the L version of the photograph.

  The second reference value is preferably 75%, more preferably 80%. The first reference value can be the number of pixels whose dimensions are Tmm when the image is printed on the L plate (8.9 cm × 12.7 cm). T can be set to 0.3 to 0.7, and more preferably 0.4 to 0.6. These reference values can be determined based on the number of pixels of the image data and the size when the image of the image data is printed.

  Further, when a target area having a partial blurring degree smaller than the first reference value exists in the photographic image, the photographic image can be determined as a focused image. With such an aspect, even when the user takes a photograph by focusing on an object that is small in the image, the photograph image can be determined to be a focused image.

  Further, when the area of the photographic image is divided into a first area near the center and a second area near the outer periphery surrounding the first area, a predetermined number of in-focus areas are included in the first area. Alternatively, when there is a predetermined ratio or more, the photographic image can be determined to be a focused image. With such an aspect, it is possible to determine that a photograph in which a person is placed in the center and the person is focused and the background is blurred is a focused image.

  That is, whether or not a photographic image is a focused image is determined based on the blurring degree of the entire photographic image data determined by a statistical method based on the partial blurring degree of each specific area in the photographic image, and the threshold. This can be done by comparing the value.

C14. Modification 14:
In the above embodiment, the image data is image data having a coefficient of each frequency obtained by discrete cosine transform (DCT). The basic pattern is also stored in the ROM 160 as coefficients obtained by discrete cosine transform (DCT). However, it is also possible to adopt a mode in which the image data has a coefficient of each frequency obtained by other orthogonal transforms. In such an aspect, it is preferable that the basic pattern is also stored as a coefficient obtained by the same orthogonal transformation. As the orthogonal transform, for example, an orthonormal transform such as a discrete cosine transform, a discrete Fourier transform, or a discrete sine transform can be applied.

C15. Modification 15:
In the above embodiment, the image data file read in step S10 of FIG. 3 is a JPEG format image data file. However, it is also possible to read an image data file of another format such as a BMP format image data file. In such an aspect, after calculating coefficients of a plurality of frequency components representing a change pattern along a predetermined direction of a target area that is a partial area of the image in advance, the processing of step S20 and subsequent steps may be performed. preferable.

C16. Modification 16:
In this embodiment, the printer 1 has a card slot 70 for inserting a memory card and reading data. However, the printer 1 may include other means for receiving the image data file instead of the card slot 70. For example, the printer 1 may include a communication unit that can receive an image data file from the outside through a wired or wireless communication line.

C17. Modification 17:
In the above embodiment, an example of the integrated printer 1 has been described (see FIG. 1). However, the present invention can be applied to various processes in which in-focus determination is performed on a plurality of image data files and image data file candidates for the in-focus image are selected. The present invention can also be realized as a computer capable of performing processing for determining whether an image of image data is blurred. In addition, information indicating whether or not the image is a blurred image determined through such processing can be generated as data and added to the image data file.

  One embodiment of the present invention can also be realized as a computer system including a display unit such as a liquid crystal display or a plasma display, an input unit such as a keyboard or a mouse, a CPU, and a main memory. In such an aspect, the CPU of the computer system has the same function as the CPU 100 of the above embodiment. Then, the image of the selected image data file is displayed on the display unit.

  Another aspect of the present invention may be a system including a projector system that projects an image data file selected as a candidate onto a plane or space. In other words, the output of the image data file is not limited to printing on a print medium, but can be in various forms such as display on a display device or external projection by a projection unit.

  In each of the above systems, after selecting an image data file candidate to be output based on the blurring degree, and further performing processing for selecting an output target from those candidates, printing, display, etc. It can also be set as the aspect which outputs. Further, for example, after selecting image data file candidates to be output from a plurality of image data files based on the degree of blur, whether or not to display these images on the display 40 and print them. The user can be prompted to make a decision.

C18. Modification 18:
In the second embodiment, the difference between the maximum value and the minimum value of the luminance is calculated based on the DCT coefficient and the signed normalization coefficient of the basic pattern (see Expression (10), Expression (7), etc.). However, the difference between the maximum value and the minimum value of the color density and luminance can also be calculated by calculating the color density and luminance at each pixel position from the DCT coefficient.

C19. Modification 19
In the above-described embodiment, the plurality of coefficients obtained by normalization or signed normalization are coefficients whose total absolute value is 1. However, normalization and signed normalization are not limited to such an aspect, but may be other aspects. That is, normalization and signed normalization can be performed so that the sum of the absolute values of a plurality of obtained coefficients becomes a constant value for each basic pattern and the luminance change pattern of the target region.

C20. Modification 20:
In the second embodiment, the sum Sja, Sjb, Sjc, Sjd of the absolute values of the AC components of the basic patterns Pja, Pjb, Pjc, Pjd is stored in the ROM 160 in advance. However, the values stored in the ROM 160 may be other values related to the sum Sja, Sjb, Sjc, Sjd of the absolute values of the AC components of the basic patterns Pja, Pjb, Pjc, Pjd.

  For example, the values stored in the ROM 160 are the values Snja, Snjb, Snjc obtained by multiplying the sum Sja, Sjb, Sjc, Sjd of the absolute values of the AC components of the basic patterns Pja, Pjb, Pjc, Pjd by √2 and dividing by UL. , Snjd. In such a case, by using Snjx (x is a, b, c or d) stored in the ROM 160, the luminance difference Ld can be obtained by the following equation.

  Ld = Sfh / Snjx (14)

  Further, in the aspect in which the magnitude of the luminance difference of the basic pattern differs for each basic pattern as in the first embodiment (see FIG. 7), the sum of the absolute values of the AC components of each basic pattern is multiplied by √2, A value divided by the luminance difference of the basic pattern (difference between the maximum value and the minimum value of the luminance) can also be stored in the ROM 160. Even if such a value is used, the luminance difference Ld can be calculated based on the sum Sfh of absolute values of DCT coefficients of the AC components in the horizontal direction of the target region.

  In any aspect, the luminance difference Ld corresponding to the difference between the maximum value and the minimum value of the luminance of the target region is expressed as “the sum of absolute values of DCT coefficients of AC components in the horizontal direction (or vertical direction) of the target region Sfh ( Or, the value obtained by dividing "Sfv)" by "the sum of the absolute values of the AC components of the basic pattern" is multiplied by a predetermined constant.

C21. Modification 21:
In the above embodiment, a part of the configuration realized by hardware may be replaced with software, and conversely, a part of the configuration realized by software may be replaced by hardware. For example, a part of the function of the CPU can be executed by a hardware circuit.

  A computer program for realizing such a function is provided in a form recorded on a computer-readable recording medium such as a floppy disk or a CD-ROM. The host computer reads the computer program from the recording medium and transfers it to the internal storage device or the external storage device. Alternatively, the computer program may be supplied from the program supply device to the host computer via a communication path. When realizing the function of the computer program, the computer program stored in the internal storage device is executed by the microprocessor of the host computer. Further, the host computer may directly execute the computer program recorded on the recording medium.

In this specification, the computer is a concept including a hardware device and an operation system, and means a hardware device that operates under the control of the operation system. The computer program causes such a computer to realize the functions of the above-described units. Note that some of the functions described above may be realized by an operation system instead of an application program.

  In this specification, the “computer-readable recording medium” is not limited to a portable recording medium such as a flexible disk or a CD-ROM, but may be an internal storage device in a computer such as various RAMs and ROMs. In addition, an external storage device fixed to a computer such as a hard disk is also included.

1 is a perspective view showing a printer 1 that is an embodiment of the present invention. FIG. The top view which expands and shows the operation panel provided with the liquid crystal display 40 and the button group 50. FIG. The flowchart which shows the processing content of "automatic printing". The figure which shows the DCT coefficient about the image of the object area | region in which the image data of JPEG format is stored. The flowchart which shows the processing content of edge pattern matching. The figure which shows the luminance level L1i of the basic pattern P01, the one-dimensional DCT coefficient Cf1i, and the normalization DCT coefficient Cfn1i (i = 1-7). The figure which shows 16 basic patterns Pj (j is 01-07, 11-15, 21-23, 31) of a luminance change. The table | surface which shows the relationship between the direction of the pattern of a luminance change, and the code | symbol of a DCT coefficient. The flowchart which shows the content of the process performed when the determination result of FIG.5 S330 is No. The flowchart which shows the content of the connection process of the edge in step S30 of FIG. The figure which shows the basic pattern group PGa containing the basic pattern of 16 brightness | luminance changes The figure which shows basic pattern group PGb containing the basic pattern of 16 brightness | luminance changes The figure which shows basic pattern group PGc containing the basic pattern of 16 brightness | luminance changes The figure which shows basic pattern group PGd containing the basic pattern of 16 brightness | luminance changes The flowchart which shows the processing content of the edge pattern matching in 2nd Example. The flowchart which shows the processing content of the edge pattern matching in 2nd Example. The table | surface which shows the selection method of the basic pattern group in step S315. The flowchart which shows the content of the connection process of the edge in 2nd Example. The flowchart which shows the content of the connection process of the edge in 2nd Example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Printer 20 ... Auto seed feeder 30 ... Paper discharge tray 40 ... Liquid crystal display 50 ... Button group 52 ... Power button 54 ... Mode button group 56 ... Setting button 58 ... Menu button group 60 ... Start button group 62 ... Stop button 70 ... Card slot 100 ... CPU
150 ... Main memory 160 ... ROM
582 ... OK button 584 ... Button 586 ... Selection button group 588 ... Print number button group 601 ... Color printing button 602 ... Monochrome printing button A1 ... Arrow indicating insertion of memory card MC into card slot 70 M11 ... "Print all" Menu display M12 ... Menu display of "Select printing" M13 ... Menu display of "Random printing" M21 ... Explanation of "Random printing" Atp ... Target area currently processed Atf ... Target area processed immediately before Bb ... Partial blurring degree of target area Bbh ... Horizontal partial blurring degree of target area Bbv ... Vertical partial blurring degree of target area Cf1i ... One-dimensional DCT coefficient of basic pattern P01 (i = 0 to 7)
Cfn1i: Normalized DCT coefficient of basic pattern P01 F00: DCT coefficient (DC component) of JPEG image data file
F01 to F77: JPEG DCT coefficient (AC component)
Fr0i: Normalization coefficient of target region (i = 0 to 7)
L1i: Luminance level at each pixel position of the basic pattern P01 (i = 0 to 7)
MC: memory card P01a-P01d: luminance change pattern obtained by inverting the basic pattern P01 vertically and horizontally P01-P31 ... basic pattern PGa-PGd ... basic pattern group Pf ... pattern of luminance change of the target area Atf processed immediately before Pj, Pja, Pjb, Pjc, Pjd ... basic pattern (j = 01 to 07, 11 to 15, 21 to 23, 31)
Pmat: Basic pattern Pp: Brightness change pattern of target region Atp currently processed S1: Frame indicating that menu is selected SDj: Sum of absolute values of differences of normalized DCT coefficient Cfnji SDsj: Signed normal Sum of absolute values of differences of generalized DCT coefficients Cfsji UL ... Unit luminance difference

Claims (24)

A printing apparatus that selects and prints image data candidates from a plurality of image data,
A data acquisition unit that acquires one image data from a plurality of image data, the image data including a plurality of coefficients representing a pattern of color change of an image along a predetermined direction, The coefficient is a coefficient corresponding to each different frequency component, a data acquisition unit,
A basic pattern memory storing a plurality of basic pattern data each including a plurality of basic coefficients representing a basic pattern of color change, wherein the plurality of basic coefficients are coefficients corresponding to different frequency components; The basic pattern data includes basic pattern memory including data representing basic blurring degree representing blurring degree in a color boundary portion;
Based on the comparison between the plurality of coefficients included in the acquired image data and the plurality of basic coefficients respectively included in the plurality of basic pattern data stored in the basic pattern memory. A basic pattern selection unit for selecting one selected basic pattern data from the pattern data;
A candidate that calculates the blur level of the image data based on the basic blur level associated with the selected basic pattern data and determines whether or not the acquired image data is a candidate based on the blur level A decision unit;
A printing unit that prints the image data selected as the candidate. The printing apparatus according to claim 1,
The basic pattern selection unit, when a predetermined first condition is satisfied, an image color change pattern along the predetermined direction in a target area that is a partial area of the image of the image data Selecting the selected basic pattern data based on a comparison of the plurality of coefficients representing the plurality of basic coefficients included in each of the plurality of basic pattern data stored in the basic pattern memory,
The candidate determination unit
Determining a partial blur level of the target area based on the basic blur level of the selected basic pattern data;
The printing apparatus which determines whether the acquired image data is made the candidate based on the partial blurring degree of the plurality of target regions included in the image of the acquired image data. The printing apparatus according to claim 2,
The basic coefficient is a coefficient corresponding to an AC component among the frequency components,
The basic pattern selection unit is configured to perform the selection based on a comparison between a coefficient corresponding to an AC component among the plurality of coefficients in the target region and the plurality of basic coefficients respectively included in the plurality of basic pattern data. Printing device that selects basic pattern data. The printing apparatus according to claim 3,
The plurality of basic coefficients are coefficients that are all positive numbers and have a constant total value,
The basic pattern selection unit further calculates, for the target region, the coefficients corresponding to the frequency components so that the total is the constant number based on a ratio of absolute values of the frequency components. A printing device. The printing apparatus according to claim 4, wherein
The plurality of basic pattern data includes data of a predetermined order,
The basic pattern selection unit, when selecting the selected basic pattern data, sequentially compares the plurality of basic coefficients of the plurality of basic pattern data with the plurality of coefficients of the target region according to the order, A printing apparatus that selects, as the selected basic pattern data, an initial basic pattern in which a difference between a plurality of basic coefficients and the plurality of coefficients is smaller than a predetermined reference. The printing apparatus according to claim 4, wherein
The plurality of basic coefficients represent a luminance change pattern in which luminance changes from a first value to a second value within a predetermined section as a basic pattern of the color change,
The basic blur level is a value corresponding to a width of a section in which luminance changes from the first value to the second value in the predetermined section,
In the first selected basic pattern data selected based on the first target area, the section in which the luminance changes, and the second target area adjacent to the first target area side by side in the predetermined direction In the second selected basic pattern data selected based on the above, the section in which the brightness changes is located at the end of the other target area in the predetermined section, and the direction of the change in the brightness Are the same, the candidate determination unit determines the partial blurring degree of at least one of the first and second target regions as the basic blurring degree of the first and second selected basic pattern data. Printing device that calculates based on the sum of The printing apparatus according to claim 6,
In the first and second selected basic pattern data, the basic pattern selection unit further has a width in which the brightness changes within a predetermined value from an end on the other target area side of the predetermined section. And the partial blurring degree of at least one of the first and second target areas is set to the first and second positions when the brightness change direction is the same. A printing apparatus that calculates based on a sum of the basic blur levels of the selected basic pattern data of a target area. The printing apparatus according to claim 6,
The printing apparatus, wherein the predetermined sections of the plurality of basic patterns have the same size. The printing apparatus according to claim 6,
The printing apparatus, wherein the plurality of basic patterns include basic patterns having different predetermined section sizes. The printing apparatus according to claim 2,
The first condition is that the sum of absolute values of the plurality of coefficients in the target area is larger than a predetermined reference value. A printing apparatus according to claim 10, wherein
The basic pattern selection unit is further based on a difference in brightness between two target regions adjacent to each other when the first condition is not satisfied and the predetermined second condition is satisfied. A printing apparatus for selecting the selected basic pattern data. The printing apparatus according to claim 2,
When the candidate determination unit determines the degree of partial blur,
Selecting the selected basic pattern data based on the plurality of coefficients representing a color change pattern of the image of the target area along the horizontal direction of the image as one of the predetermined directions; Calculate the horizontal blur of the target area based on the basic blur of the pattern data,
Selecting the selected basic pattern data based on the plurality of coefficients representing a color change pattern of the image of the target area along the vertical direction of the image as one of the predetermined directions; Calculate the vertical blur of the target area based on the basic blur of the pattern data,
A printing apparatus that determines the partial blurring degree of the target region based on the horizontal blurring degree and the vertical blurring degree. The printing apparatus according to claim 12, wherein
The candidate determination unit determines the partial blurring degree of the target region based on a larger one of the horizontal blurring degree and the vertical blurring degree. The printing apparatus according to claim 2,
The candidate determination unit determines a target area that is a predetermined size included in the image of the acquired image data and has a partial blurring degree smaller than a first reference value as a second reference value. A printing apparatus that uses the acquired image data as the candidate when there is a region that includes more than a region. The printing apparatus according to claim 14, wherein
The printing apparatus according to claim 1, wherein the first and second reference values are determined based on the number of pixels of the image of the image data and a size of the image data when the image is printed. The printing apparatus according to claim 1,
The plurality of basic pattern data is classified into a plurality of basic pattern data groups each including one or more basic pattern data,
The basic pattern selection unit
A group selector that selects one of the plurality of basic pattern data groups based on a part of the plurality of coefficients of the image data when selecting the selected basic pattern data;
Based on a comparison between the plurality of coefficients included in the image data and the plurality of basic coefficients included in the basic pattern data of the selected basic pattern data group, one of the plurality of basic pattern data is selected. A printing device that selects two basic pattern data. The printing apparatus according to claim 16, wherein
The basic pattern selection unit, when a predetermined first condition is satisfied, an image color change pattern along the predetermined direction in a target area that is a partial area of the image of the image data Selecting the selected basic pattern data based on a comparison of the plurality of coefficients representing the plurality of basic coefficients included in each of the plurality of basic pattern data stored in the basic pattern memory,
The candidate determination unit
Determining a partial blur level of the target area based on the basic blur level of the selected basic pattern data;
Based on the partial blurring degree of a plurality of the target areas included in the image of the acquired image data, it is determined whether or not the acquired image data is the candidate (old 2)
The basic coefficient is a coefficient corresponding to an AC component among the frequency components,
The basic pattern selection unit is configured to perform the selection based on a comparison between a coefficient corresponding to an AC component among the plurality of coefficients in the target region and the plurality of basic coefficients respectively included in the plurality of basic pattern data. Select basic pattern data (old 3)
The plurality of basic coefficients are coefficients whose sum of absolute values is a constant value,
The basic pattern selection unit further calculates, based on the frequency components, the coefficients corresponding to the frequency components so that the sum of absolute values becomes the constant value for the target region. The printing apparatus according to claim 16, wherein
The group selection unit is configured to select one of the plurality of basic pattern data groups based on a sign of a coefficient corresponding to the first and second low frequency AC components among the frequency components. The printer that selects one. The printing apparatus according to claim 1,
The basic pattern selection unit, when a predetermined first condition is satisfied, an image color change pattern along the predetermined direction in a target area that is a partial area of the image of the image data The selected basic pattern in association with the target area based on a comparison of the plurality of coefficients representing the plurality of basic coefficients respectively included in the plurality of basic pattern data stored in the basic pattern memory Select the data
The candidate determination unit
Based on the basic blur level of the selected basic pattern data, determine a partial blur level of the target area included in the image of the acquired image data,
Determining a first evaluation value corresponding to a difference between a maximum value and a minimum value of color density in the target region based on the plurality of coefficients of the target region;
Determining a second evaluation value related to a plurality of target areas included in the image of the acquired image data based on the first evaluation value of each target area included in the plurality of target areas;
Whether the acquired image data is the candidate based on the partial blurring degree of each target area included in the plurality of target areas when the second evaluation value satisfies a predetermined second condition A printing device that determines whether or not. The printing apparatus according to claim 19, wherein
The printing apparatus, wherein the plurality of target areas are a plurality of target areas that exist continuously along the predetermined direction and satisfy a predetermined condition regarding the selected basic pattern data. The printing apparatus according to claim 20, wherein
The predetermined condition relating to the selected basic pattern data includes a condition that the basic pattern represented by the selected basic pattern data includes a portion where luminance changes at at least one end of the basic pattern. Printing device included as part. The printing apparatus according to claim 19, wherein
The candidate determination unit is the basic coefficient of the selected basic pattern data, which is the sum of the absolute values of the coefficients corresponding to the alternating current component among the frequency components and the coefficient of the target region, and out of the frequency components A printing apparatus that obtains the first evaluation value based on a ratio of a coefficient corresponding to an AC component to a sum of absolute values. A method for calculating the degree of image blur of image data,
(A) Step of obtaining image data including a plurality of coefficients representing a pattern of color change of an image along a predetermined direction, wherein the plurality of coefficients are coefficients corresponding to different frequency components. When,
(B) preparing a plurality of basic pattern data each including a plurality of basic coefficients representing a basic pattern of color change, wherein the plurality of basic coefficients are coefficients corresponding to different frequency components, The basic pattern data includes data representing basic blurring degree representing blurring degree in a color boundary part,
(C) selecting one of the plurality of basic pattern data based on a comparison between the plurality of coefficients included in the image data and the plurality of basic coefficients respectively included in the plurality of basic pattern data A process of selecting basic pattern data;
(D) calculating an image blur level of the image data based on the basic blur level associated with the selected basic pattern data. A computer program for causing a computer to execute a process of calculating the degree of blur of an image of image data,
A function of acquiring image data including a plurality of coefficients representing a pattern of color change of an image along a predetermined direction, wherein the plurality of coefficients are coefficients corresponding to different frequency components;
A function of preparing a plurality of basic pattern data each including a plurality of basic coefficients representing a basic pattern of color change, wherein the plurality of basic coefficients are coefficients corresponding to different frequency components; Includes data representing the basic blur level representing the blur level at the color boundary, and
Based on a comparison between the plurality of coefficients included in the image data and the plurality of basic coefficients included in the plurality of basic pattern data, one selected basic pattern data from the plurality of basic pattern data With the ability to select
A function of calculating the blur level of the image data based on the basic blur level associated with the selected basic pattern data;
A computer program for causing the computer to realize the above.

Images