JP6291546B2 - Image evaluation apparatus, image evaluation method, and program - Google Patents

Description

  The disclosed technology relates to an image evaluation apparatus, an image evaluation method, and a program.

  A part of the printed matter printed by the printing machine may be a defective printed matter. In view of this, there has been proposed an inspection method for determining the quality of a printed material by determining whether there is a defective printed material that cannot be overlooked based on an image obtained by reading the printed material with a scanner (for example, Patent Document 1 and Patent Document 2). See).

  In Patent Document 1, a density difference between the output master image and the scan image to be inspected is calculated, and by determining whether the calculated density difference is acceptable, defects included in the scan image to be inspected are overlooked. An inspection method for determining whether or not it is possible is disclosed.

  Patent Document 2 discloses a method for detecting a flat panel display device using a visual model. This method is a method of recognizing a captured image of a test panel and a reference image simulated by a background by a difference recognition generation system and evaluating the quality of a display device using a visual model.

Japanese Patent No. 4407588 JP 2006-139777 A

  By the way, an image (inspection image) obtained by scanning an image included in a printed material for inspection with a scanner, for example, may have a streak defect (hereinafter referred to as “streaks defect”). The streak defect occurs due to, for example, a malfunction of the printing press (for example, a malfunction in the structure). In addition, many streak defects caused by a malfunction of the printing press tend to appear in a streak pattern along a specific direction. Such streak defects often have a degree of defects exceeding an allowable range.

  However, none of the methods described in Patent Document 1 and Patent Document 2 can accurately determine whether or not a streak defect caused by a structural failure of the printing press is an unacceptable streak defect. There was a problem.

  The disclosed technology has been proposed in view of such circumstances, and an image evaluation apparatus and an image that can determine with high accuracy whether or not an image having unacceptable streak defects is a printed matter. An object is to provide an evaluation method and an image evaluation program.

  To achieve the above object, the image evaluation apparatus according to the first aspect of the present invention stores a sample image information indicating a virtual sample image assuming a sample image included in a sample printed material recognized as a non-defective product. And a reading unit that reads an inspection object image included in an inspection object printed matter obtained by printing an image indicated by the image information on a recording medium by a printing device based on image information indicating an image corresponding to a sample image. Obtained by reading the inspection target image with the reading unit that reads the inspection target image so that the resolution in the specific direction is lower than the resolution in the direction intersecting the specific direction, and the sample image information stored in the storage unit and the reading unit Based on the difference value between the inspection target image information indicating the inspection target image and the image corresponding to the sample image, a streak in a specific direction is obtained from the inspection target image indicated by the inspection target image information. An extraction unit that extracts streak-like defects having a shape, and a resolution in a direction intersecting a specific direction with a resolution in a specific direction of the inspection target image indicated by the inspection target image information obtained by reading the inspection target image by the reading unit And extracting the streak defect from the inspection target image indicated by the inspection target image information based on the difference value between the sample image information and the inspection target image information, and the streak extracted by the extraction unit And an evaluation unit that evaluates the visibility of the defect using a mask effect visual model. As a result, it is possible to determine with high accuracy whether or not the image has an unacceptable streak defect as compared with the case without the present configuration. Moreover, the evaluation accuracy of the visibility of the streak defect can be improved as compared with the case where this configuration is not provided. Further, compared to the case without this configuration, it is possible to suppress a decrease in evaluation accuracy while reducing the load on the reading process. Furthermore, it is possible to extract a streak defect in a specific direction with higher accuracy than in the case where this configuration is not provided.

  In the image evaluation apparatus according to the first aspect of the present invention, as in the invention according to the second aspect of the present invention, the evaluation unit uses the mask effect visual model for a specific direction to check the streak defect visibility. It is good also as what evaluates.

  As in the invention according to the fourth aspect of the present invention, the image evaluation apparatus according to any one of the first aspect or the second aspect of the present invention has a streak defect to be evaluated by the evaluation unit. Alternatively, a spatial frequency streak defect corresponding to human visual characteristics may be used. Thereby, compared with the case where it does not have this structure, the evaluation accuracy of the visibility of a streak defect can be improved.

  As in the invention according to the fourth aspect of the present invention, the image evaluation apparatus according to the first aspect or the second aspect of the present invention is configured so that the streak defect to be evaluated by the evaluation unit is a human visual characteristic. It may be a streak defect in the color space corresponding to. Thereby, compared with the case where it does not have this structure, the evaluation accuracy of the visibility of a streak defect can be improved.

  In the image evaluation device according to any one of the first to fourth aspects of the present invention, as in the invention according to the fifth aspect of the present invention, the extraction unit is inspected by the inspection target image information. The streak defect may be extracted by targeting a predetermined region in the target image. Thereby, compared with the case where this configuration is not provided, the processing load can be reduced.

  The image evaluation apparatus according to the fifth aspect of the present invention is a region that is preliminarily assumed as a region where a streak defect is formed, as in the invention according to the sixth aspect of the present invention. It is good also as what. Thereby, compared with the case where it does not have this structure, the fall of evaluation accuracy can be suppressed, reducing the load concerning a process.

  As in the invention according to the seventh aspect of the present invention, the image evaluation apparatus according to the sixth aspect of the present invention further includes a predetermined region, among the inspection target images indicated by the inspection target image information. An area corresponding to a difference value exceeding a threshold value may be used as a difference value between the inspection target image and the sample image indicated by the sample image information. Thereby, compared with the case where it does not have this structure, the fall of evaluation accuracy can be suppressed further, reducing the load concerning a process.

  As in the invention according to the eighth aspect of the present invention, the image evaluation apparatus according to the sixth aspect or the seventh aspect of the present invention is configured so that an area assumed in advance is at least one of the printing device and the reading unit. It may be specified based on characteristics. Thereby, compared with the case where it does not have this structure, the streak defect of a specific direction can be extracted with high precision with a simple structure.

  In the image evaluation apparatus according to the eighth aspect of the present invention, the characteristics of the printing apparatus may include the characteristics of deterioration of components of the printing apparatus over time as in the invention according to the ninth aspect of the present invention. Good. Thereby, compared with the case where it does not have this structure, the streak defect of a specific direction can be extracted still more accurately.

  As in the invention according to the tenth aspect of the present invention, the image evaluation apparatus according to the eighth aspect or the ninth aspect of the present invention is characterized in that the characteristics of the reading unit are the deterioration characteristics of the constituent members in the reading unit over time. May be included. Thereby, compared with the case where it does not have this structure, the streak defect of a specific direction can be extracted still more accurately.

  As in the invention according to the eleventh aspect of the present invention, the image evaluation apparatus according to the fifth aspect of the present invention is configured so that the predetermined region is the inspection of the inspection target image indicated by the inspection target image information. It is good also as a field corresponding to a difference value exceeding a threshold with a difference value between the target image and the sample image indicated by the sample image information. Thereby, compared with the case where it does not have this structure, the fall of evaluation accuracy can be suppressed further, reducing the load concerning a process.

  As in the invention according to the twelfth aspect of the present invention, the image evaluation apparatus according to any one of the first aspect to the eleventh aspect of the present invention sets the specific direction to the conveyance direction of the recording medium in the printing apparatus. It is good also as what. Thereby, compared with the case where this structure is not provided, the streak defect of a specific direction can be extracted with high precision.

  The image evaluation apparatus according to the twelfth aspect of the present invention may be configured such that the printing method of the printing apparatus is a single pass method, as in the thirteenth aspect of the present invention. Thereby, compared with the case where it does not have this structure, the streak defect of a specific direction can be extracted still more accurately.

  In the image evaluation apparatus according to any one of the first to eleventh aspects of the present invention, as in the invention according to the fourteenth aspect of the present invention, the printing apparatus divides the inspection target image into a plurality of lines. When recording is performed on the recording medium for each line, the specific direction may be the main scanning direction in the printing apparatus. Thereby, compared with the case where this structure is not provided, the streak defect of a specific direction can be extracted with high precision.

  As in the invention according to the fifteenth aspect of the present invention, the image evaluation apparatus according to the fourteenth aspect of the present invention may be configured such that the printing method of the printing apparatus is a shuttle scan method. Thereby, compared with the case where it does not have this structure, the streak defect of a specific direction can be extracted still more accurately.

  The image evaluation apparatus according to any one of the first to fifteenth aspects of the present invention, as in the invention according to the sixteenth aspect of the present invention, includes at least sample image information and inspection target image information. An exclusion unit that excludes the influence of the characteristics of the reading unit from the inspection target image information is further included, and the extraction unit uses the exclusion unit to influence the characteristics of the reading unit from at least the inspection target image information of the sample image information and the inspection target image information. The streak defect may be extracted from the inspection target image indicated by the inspection target image information based on the difference value between the sample image information and the inspection target image information obtained by eliminating. Thereby, compared with the case where it does not have this structure, the streak defect of a specific direction can be extracted still more accurately.

  The image evaluation apparatus according to any one of the first to sixteenth aspects of the present invention is a streak defect whose visibility has been evaluated by the evaluation unit as in the seventeenth aspect of the present invention. It is good also as what further contains the warning part which warns when it determines with it being a streak-like defect which can be visually recognized. Thereby, compared with the case where this configuration is not provided, it is possible to easily grasp that the inspection target printed matter is printed with an inspection target image having an unacceptable streak defect.

  The image evaluation apparatus according to the seventeenth aspect of the present invention is an inspection including a streak defect in which the warning unit is determined to be a visible streak defect as in the eighteenth aspect of the present invention. A warning may be issued by directly processing the target printed matter. Thereby, compared with the case where this configuration is not provided, it is possible to specify the inspection target printed matter on which the inspection target image having unacceptable streak defects is printed with a simple configuration.

  In the image evaluation apparatus according to the eighteenth aspect of the present invention, as in the nineteenth aspect of the present invention, the warning unit provides a specific support mark that supports the specification of the position of the streak defect in the printed material to be inspected. It is good also as what warns by giving to this inspection object printed matter. Thereby, compared with the case where this structure is not provided, the position of the streak defect can be easily grasped with a simple structure.

  In the image evaluation apparatus according to the nineteenth aspect of the present invention, as in the invention according to the twentieth aspect of the present invention, the warning unit specifies the specific support mark at the position where the streak defect in the printed material to be inspected extends in the specific direction. It is good also as what warns by giving. Thereby, compared with the case where it does not have this structure, the position of a stripe defect can be grasped | ascertained still more easily.

  As in the invention according to the twenty-first aspect of the present invention, the image evaluation apparatus according to the nineteenth aspect or the twentieth aspect of the present invention sets the type of the specific support mark according to the visibility of the streak defect. It may be determined as follows. Thereby, compared with the case where it does not have this structure, the magnitude | size of the visibility of a stripe defect can be grasped | ascertained easily.

  In the image evaluation device according to any one of the seventeenth to twenty-first aspects of the present invention, as in the invention according to the twenty-second aspect of the present invention, the warning unit further has visibility by the evaluation unit. When it is determined that the evaluated streak defect is a visible streak defect, a warning may be given by displaying predetermined information on the display unit. Thereby, compared with the case where this configuration is not provided, it is possible to easily grasp that the inspection target printed matter is printed with an inspection target image having an unacceptable streak defect.

  In the image evaluation device according to the seventeenth aspect of the present invention, as in the invention according to the twenty-third aspect of the present invention, the warning part is a streak in which the streak defect whose visibility is evaluated by the evaluation part can be visually recognized. If it is determined to be a defect, a warning may be given by displaying predetermined information on the display unit. Thereby, compared with the case where this configuration is not provided, it is possible to easily grasp that the inspection target printed matter is printed with an inspection target image having an unacceptable streak defect.

  In the image evaluation device according to the twenty-second aspect or the twenty-third aspect of the present invention, as in the invention according to the twenty-fourth aspect of the present invention, the predetermined information is a streak determined to be a visible streak defect. It is good also as what contains the specific information which pinpoints inspection object printed matter containing a shape defect. Thereby, compared with the case where this configuration is not provided, it is possible to specify the inspection target printed matter on which the inspection target image having unacceptable streak defects is printed with a simple configuration.

  In the image evaluation device according to any one of the twenty-second to twenty-fourth aspects of the present invention, as in the invention according to the twenty-fifth aspect of the present invention, the display unit displays an image, and the predetermined information is An inspection target image having a streak defect and a specification support image that supports the specification of the position of the streak defect in the inspection target image may be included. Thereby, compared with the case where this structure is not provided, the position of the streak defect can be easily grasped with a simple structure.

  In order to achieve the above object, an image evaluation method according to a twenty-sixth aspect of the present invention is a storage unit that stores sample image information indicating a virtual sample image assuming a sample image included in a sample printed material recognized as a non-defective product. The inspection target image included in the inspection target print obtained by printing the image indicated by the image information on the recording medium by the printing device based on the sample image information stored in the image and the image information indicating the image corresponding to the sample image An inspection target image obtained by reading an inspection target image by a reading unit that reads the inspection target image such that the resolution in a specific direction is lower than the resolution in a direction intersecting the specific direction. A streak having a streak pattern in a specific direction from the inspection target image indicated by the inspection target image information based on the difference value with the inspection target image information indicating the image corresponding to the sample image In the extraction step of extracting the depression, after the resolution in the specific direction of the inspection target image indicated by the inspection target image information obtained by reading the inspection target image by the reading unit is aligned with the resolution in the direction intersecting the specific direction, Based on the difference value between the sample image information and the inspection target image information, an extraction step for extracting the streak defect from the inspection target image indicated by the inspection target image information, and the visibility of the streak defect extracted by the extraction step And an evaluation step for evaluating using a mask effect visual model. As a result, it is possible to determine with high accuracy whether or not the image has an unacceptable streak defect as compared with the case without the present configuration. Moreover, the evaluation accuracy of the visibility of the streak defect can be improved as compared with the case where this configuration is not provided. Further, compared to the case without this configuration, it is possible to suppress a decrease in evaluation accuracy while reducing the load on the reading process. Furthermore, it is possible to extract a streak defect in a specific direction with higher accuracy than in the case where this configuration is not provided.

  In order to achieve the above object, a program according to a twenty-seventh aspect of the present invention is a computer serving as an extraction unit and an evaluation unit in an image evaluation device according to any one of the first to twenty-fifth aspects of the present invention. It is for making the function. As a result, it is possible to determine with high accuracy whether or not the image has an unacceptable streak defect as compared with the case without the present configuration.

  According to the disclosed technology, it is possible to determine with high accuracy whether or not an image having an unacceptable streak defect is a printed matter.

It is a schematic block diagram which shows an example of the whole structure of the image evaluation apparatus which concerns on 1st Embodiment. It is a functional block diagram which shows an example of the principal part function of the reading part contained in the image evaluation apparatus which concerns on 1st Embodiment. It is a functional block diagram which shows an example of the principal part function of the extraction part contained in the image evaluation apparatus which concerns on 1st Embodiment. It is a functional block diagram which shows an example of the principal part function of the evaluation part contained in the image evaluation apparatus which concerns on 1st Embodiment. It is a functional block diagram which shows an example of the principal part function of the warning part contained in the image evaluation apparatus which concerns on 1st Embodiment. It is a block diagram which shows an example of a structure of the electric system of the image evaluation apparatus which concerns on 1st Embodiment. It is a block diagram which shows an example of a structure of the electric system of the scanner contained in the image evaluation apparatus which concerns on 1st Embodiment. It is a schematic diagram which shows an example of the sorting system by the sorter contained in the image evaluation apparatus which concerns on 1st Embodiment. 1 is a schematic configuration diagram illustrating an example of a configuration of a conveyance system and its periphery in an image evaluation apparatus according to a first embodiment, and is a schematic configuration diagram illustrating an example of a conveyance mode of printed matter and an arrangement example of a scanner, a recording unit, and a sorter. It is. It is a schematic diagram which shows an example of the aspect of the inspection target printed matter in which the specific support mark is recorded by the recording unit included in the image evaluation device according to the first embodiment. FIG. 11 is a schematic diagram illustrating an example of an aspect of a printed material to be inspected in which a specific support mark is recorded by a recording unit included in the image evaluation device according to the first embodiment, which is different from the aspect illustrated in FIG. 10. FIG. 12 is a schematic diagram illustrating an example of a printed matter to be inspected in which a specific support mark is recorded by a recording unit included in the image evaluation device according to the first embodiment, which is different from the modes illustrated in FIGS. 10 and 11. is there. FIG. 13 is a schematic diagram illustrating an example of a printed matter to be inspected in which a specific support mark is recorded by a recording unit included in the image evaluation device according to the first embodiment, and is different from the modes illustrated in FIGS. 10 to 12. is there. It is a flowchart which shows an example of the flow of the image evaluation process which concerns on 1st-3rd embodiment. 6 is a flowchart illustrating an example of a flow of a reading control process included in an image evaluation process according to the first embodiment. It is a flowchart which shows an example of the flow of the sample image reading process which concerns on 1st Embodiment. It is a flowchart which shows an example of the flow of the test object image reading process which concerns on 1st Embodiment. It is a flowchart which shows an example of the flow of the extraction process included in the image evaluation process which concerns on 1st Embodiment. It is a flowchart which shows an example of the flow of the evaluation process contained in the image evaluation process which concerns on 1st Embodiment. It is a flowchart which shows an example of the flow of the warning process contained in the image evaluation process which concerns on 1st Embodiment. It is a mode figure which shows an example of the display mode by the display part contained in the image evaluation apparatus which concerns on 1st Embodiment. It is a functional block diagram which shows the modification (the 1) of the extraction part contained in the image evaluation apparatus which concerns on 1st Embodiment. It is a functional block diagram which shows the modification (the 2) of the extraction part contained in the image evaluation apparatus which concerns on 1st Embodiment. It is a functional block diagram which shows an example of the principal part function of the evaluation part contained in the image evaluation apparatus which concerns on 2nd Embodiment. It is a schematic diagram which shows an example of a structure of image DB contained in the image evaluation apparatus which concerns on 2nd Embodiment. It is a schematic diagram which shows an example of a structure of statistical model DB contained in the image evaluation apparatus which concerns on 2nd Embodiment. It is a block diagram which shows an example of a structure of the electric system of the image evaluation apparatus which concerns on 2nd Embodiment. It is a flowchart which shows an example of the flow of DB update process which concerns on 2nd Embodiment. It is a flowchart which shows an example of the flow of the evaluation process contained in the image evaluation process which concerns on 2nd Embodiment. It is a functional block diagram which shows the modification of the evaluation part contained in the image evaluation apparatus which concerns on 2nd Embodiment. It is a functional block diagram which shows an example of the principal part function of the warning part contained in the image evaluation apparatus which concerns on 3rd Embodiment. It is explanatory drawing with which it uses for description of the image quality index used by customer characteristic DB contained in the image evaluation apparatus which concerns on 3rd Embodiment. It is a conceptual diagram which shows how to obtain | require an image quality index. It is a schematic diagram which shows an example of a structure of customer characteristic DB contained in the image evaluation apparatus which concerns on 3rd Embodiment. It is a graph which shows an example of distribution of OK / NG with respect to an image quality index. It is a block diagram which shows an example of a structure of the electric system of the image evaluation apparatus which concerns on 3rd Embodiment. It is a flowchart which shows an example of the flow of the warning process contained in the image evaluation process which concerns on 3rd Embodiment. It is a schematic diagram which shows an example of the evaluation result by the evaluation part which concerns on 1st Embodiment. It is a functional block diagram which shows the modification (the 3) of the extraction part contained in the image evaluation apparatus which concerns on 1st Embodiment.

  Hereinafter, an exemplary embodiment of an image evaluation apparatus according to the disclosed technique will be described with reference to the accompanying drawings.

  [First Embodiment]

  FIG. 1 shows an example of main functions of the image evaluation apparatus 10 according to the first embodiment. As shown in FIG. 1, the image evaluation apparatus 10 is an apparatus that evaluates an image of a printed matter printed by a printing apparatus 12. When image information indicating an image is input, the printing apparatus 12 creates a printed material by printing the image indicated by the image information on a recording sheet, which is an example of a recording medium, based on the image information.

  In the first embodiment, a single-pass inkjet printer is employed as the printing apparatus 12, but the disclosed technique is not limited to this. For example, instead of a single-pass inkjet printer, a shuttle scan inkjet printer may be employed. In place of the ink jet printer, a xerographic printer, a thermal head printer, a magnetography printer, a printing press, or the like may be employed. In the first embodiment, the printing apparatus 12 employs an ink jet printer that transports recording paper by holding and rotating one or more recording papers on the outer peripheral surface of the drum. The printing apparatus 12 has a recording head that ejects ink droplets, and the recording paper is recorded by passing the recording paper so that the recording surface of the recording paper faces the ink ejection port of the recording head. This is a printer of a system for recording an image by ejecting ink onto a surface.

  The printing apparatus 12 takes in a recording sheet from the outside and conveys the loaded recording sheet in a specific direction. Printing is performed by recording the image indicated by the image information on the recording paper by ejecting ink droplets from the recording head based on the input image information onto the recording paper conveyed in a specific direction. Do. In this manner, the printed material on which the image is printed on the recording paper is discharged to a predetermined area (for example, a paper discharge tray). Here, the “specific direction” refers to, for example, a sub-scanning direction that intersects the main scanning direction. Here, in order to avoid complications, a direction orthogonal to the main scanning direction is adopted as an example of the “specific direction”.

  The printed matter printed by the printing apparatus 12 is roughly classified into a sample printed matter 14 and an inspection subject printed matter 16 that has a print content corresponding to the printed content of the sample printed matter 14 and is a target to be compared with the sample printed matter 14. The The “printing content” here does not include defects in the image formed in the printing process. FIG. 1 illustrates a single sample print 14 and a single inspection target print 16 having a print content corresponding to the print content of the sample print 14 in order to avoid complications. The disclosed technique is not limited to this. For example, a plurality of sample prints 14 and inspection object prints 16 may be provided. However, in the image evaluation apparatus 10, it is assumed that the images are compared with respect to the sample printed matter 14 and the inspection target printed matter 16 whose print contents are not different from each other.

  The sample printed matter 14 refers to a printed matter recognized as a non-defective product, for example. The printed matter recognized as a non-defective product is, for example, a printable product that can be shipped and printed with an image that is recognized as having good visibility (for example, good image quality). In the first embodiment, in order to avoid complications, an image that has been recognized by the image evaluation apparatus 10 in the past as having been confirmed to have good visibility is employed as the “image that has been confirmed to have good visibility”. However, the disclosed technique is not limited to this. For example, an image that the inspector determines with an unaided eye, an image having no defect or an image in which a defect is acceptable according to a predetermined standard may be adopted. Hereinafter, an image printed on the sample printed matter 14 is referred to as a sample image. In the first embodiment, in order to avoid complications, the image printed on the sample print 14 is used as a sample image. However, the present invention is not limited to this, and a part of the image printed on the sample print 14 is used as a sample image. It is good.

  On the other hand, the inspection target printed matter 16 is a printed matter in which an image (hereinafter referred to as “inspection target image”) is printed on a recording sheet by the printing apparatus 12 based on image information indicating an image corresponding to a sample image, for example. Point to. In the first embodiment, in order to avoid complications, the image printed on the inspection target print 16 is used as the inspection target image. However, the present invention is not limited to this, and a part of the image printed on the inspection target print 16 is used. May be the image to be inspected.

  By the way, there is known an image evaluation apparatus that evaluates whether or not an excellent printed matter (inspection target printed matter 16 including an inspection target image in which streak defects cannot be visually recognized) from a simple density difference value between the sample image and the inspection target image. ing. However, in a conventional image evaluation apparatus, it is difficult to determine whether or not an acceptable streak defect depends on the degree of streak defect (for example, a minute streak defect). For this reason, the evaluation accuracy of the printed material 16 to be inspected decreases depending on the degree of the streak defect. Therefore, in order to solve such a problem, the image evaluation apparatus 10 according to the first embodiment includes a reading unit 18, an extraction unit 20, an evaluation unit 22, and a warning unit 24.

  The reading unit 18 reads a sample image included in the sample printed material 14 and an inspection target image included in the inspection target printed material 16.

  The extraction unit 20 includes sample image information obtained by reading the sample image by the reading unit 18 and inspection target image information that is an inspection target image obtained by reading the inspection target image by the reading unit 18 and corresponds to the sample image. The difference value is calculated. Then, based on the calculated difference value, a streak defect having a streak pattern in a specific direction (for example, a pattern extending linearly in a specific direction) from the inspection target image indicated by the inspection target image information (this first implementation) In the form, as an example, a streak defect image that is an image showing a streak defect) is extracted.

  The evaluation unit 22 evaluates the visibility of the line defect. Here, the “visibility of the streak defect” refers to the visibility of the streak defect specified by the streak defect image extracted by the extraction unit 20, for example. Further, “evaluating visibility” here refers to, for example, deriving the visual strength of a line defect. In the first embodiment, an example in which the evaluation unit 22 derives the visual strength of the streak defect will be described. However, the present invention is not limited to this. For example, the derived visual strength can be determined by a specific human vision. The evaluation unit 22 may be assigned a determination function for determining whether or not. Here, the “specific human” refers to a person who observes the inspection object 16, for example. Further, the “determination function” here refers to a function corresponding to a function of a warning necessity determination unit 44 described later, for example.

  The warning unit 24 warns when it is determined that the streak defect whose visibility is evaluated by the evaluation unit 22 is a visible streak defect. The “visible streak defect” mentioned here refers to a streak defect that is visible to a specific person, for example.

  FIG. 2 shows an example of the main functions of the reading unit 18. The reading unit 18 includes a reading execution unit 23 and an exclusion unit 25. The reading execution unit 23 performs reading of the sample image included in the sample printed matter 14 and reading of the inspection target image included in the inspection target printed matter 16 to acquire sample image information and inspection target image information.

  Here, the reading execution unit 23 reads the sample image so that the resolution in the specific direction and the resolution in the direction crossing the specific direction are aligned, and the resolution in the specific direction is lower than the resolution in the direction crossing the specific direction. Read the image to be inspected. That is, in the first embodiment, the reading execution unit 23 reads the sample image so that the resolution in the specific direction and the resolution in the main scanning direction are aligned, and the resolution in the specific direction is lower than the resolution in the main scanning direction. Read the image to be inspected. As a result, the resolution (vertical / horizontal resolution) in the specific direction and main scanning direction of the sample image differs from the vertical / horizontal resolution of the inspection target image. Here, as an example of the resolution in the main scanning direction when reading the inspection target image, a resolution that is different from the resolution in the specific direction and the main scanning direction when reading the sample image included in the sample print 14 is adopted. Yes.

  In the first embodiment, as described above, a case where the vertical and horizontal resolution of the sample image is different from the vertical and horizontal resolution of the inspection target image is illustrated, but the disclosed technique is not limited to this. For example, The sample image and the inspection target image may have the same vertical / horizontal resolution, and the resolution in the specific direction and the resolution in the main scanning direction may also be the same. In this case, the resolution in the specific direction and the resolution in the main scanning direction may be the same for each of the sample image and the inspection target image. In this case, the later-described resolution adjustment unit 26 shown in FIG. 3 is not necessary.

  Further, the vertical and horizontal resolutions of the sample image and the inspection target image do not need to match, and the resolution in the specific direction and the resolution in the main scanning direction do not have to match. In this case, although the later-described resolution adjustment unit 26 shown in FIG. 3 is not necessary, for example, an extraction unit 20C shown in FIG. 39 is applied instead of the extraction unit 20 shown in FIG. In this case, as an example, a resolution adjusting unit 26A described later is required before the evaluation unit 22, as shown in FIG. As an example in which the vertical and horizontal resolutions of the sample image and the inspection target image are identical and the resolution in the specific direction and the resolution in the main scanning direction are not identical as described above, a line sensor is used as the reading unit 18. The case where it uses is mentioned.

  The exclusion unit 25 excludes the influence of the characteristics of the reading unit 18 from at least the inspection target image information of the sample image information and the inspection target image information acquired by the reading execution unit 23. The “characteristic” mentioned here does not include the characteristic that the resolution in the characteristic direction is different from the resolution in the main scanning direction. Here, in order to avoid complications, an example in which the exclusion unit 25 excludes the influence of the characteristics of the reading execution unit 23 from the sample image information and the inspection target image information acquired by the reading execution unit 23 will be described. However, the disclosed technique is not limited to this. For example, the influence of the characteristic of the reading execution unit 23 may be excluded from the inspection target image information without removing the influence of the characteristic of the reading execution unit 23 from the sample image information. Further, for example, at least the inspection target image information of the sample image information and the inspection target image information may be excluded from the influence of the characteristic of the reading unit 23 as well as the influence of the characteristic of the exclusion unit 25. Also. The influence of the characteristic of the reading unit 23 may be ignored and the influence of the characteristic of the exclusion unit 25 may be excluded.

  FIG. 3 shows an example of the main functions of the extraction unit 20. The extraction unit 20 includes a resolution adjustment unit 26, a subtraction unit 28, a streak region cutout unit 30, and a specific direction image extraction unit 32. The resolution adjustment unit 26 adjusts the resolution of the inspection target image indicated by the inspection target image information from which the influence of the characteristics of the reading unit 18 is excluded by the exclusion unit 24. In the first embodiment, for example, the resolution adjustment unit 26 aligns the inspection target image indicated by the inspection target image information in which the influence of the characteristics of the reading execution unit 23 is excluded by the exclusion unit 24 and then the inspection target. The resolution in the specific direction of the image is set to the resolution in the main scanning direction. Note that “alignment” here refers to processing for correcting skew, deviation in the transport direction, and the like.

  The subtracting unit 28 calculates a difference value between the inspection target image information indicating the inspection target image whose resolution is adjusted by the resolution adjusting unit 26 and the sample image information from which the influence of the characteristics of the reading unit 18 is excluded by the removing unit 24. . In this case, for example, the subtracting unit 28 compares the inspection target image whose resolution is adjusted by the resolution adjusting unit 26 and the sample image indicated by the sample image information from which the influence of the characteristics of the reading performing unit 23 is excluded by the removing unit 24. A difference value for the density is calculated.

  The streak region cutout unit 30 targets a predetermined region of the inspection target image whose resolution is adjusted by the resolution adjustment unit 26, and based on the difference value calculated by the subtraction unit 28, A region having a pattern (hereinafter referred to as “striped region”) is cut out. The “striated pattern” referred to here includes a streak pattern in a direction other than the specific direction in addition to the streak pattern in a specific direction. In addition, the “predetermined region” referred to here refers to a region assumed in advance as a region where a streak pattern is formed, for example. In the first embodiment, as an example of “an area assumed in advance as an area on which a streak pattern is formed”, a structural defect portion of the printing apparatus 12, an area with a large ink discharge amount, and the reading unit 18. An area specified based on characteristics (characteristics of the reading execution unit 23 as an example) is employed. Here, as an example, a region corresponding to each of a structural defect portion of the printing apparatus 12, a region having a large ink discharge amount, and a structural defect portion of the reading execution unit 23 is referred to as “region where a streak pattern is formed. It is adopted as an area assumed in advance.

  As described above, in the first embodiment, the “area assumed in advance as the area where the streak pattern is formed” is specified based on the characteristics of the printing apparatus 12 and the characteristics of the reading unit 18. The disclosed technique is not limited to this. For example, the “area that is assumed in advance as the area where the streak pattern is formed” may be specified based on the characteristics of the printing device 12 or the characteristics of the reading unit 18. In addition, the characteristics of the printing apparatus 12 may include deterioration characteristics over time in the constituent members of the printing apparatus 12. For example, among the structural members of the printing apparatus 12, a structural member that is expected to deteriorate with time and that has a characteristic of deterioration with time of the structural member that is expected to affect image defects in advance is described as a “streaked pattern”. You may specify the area | region assumed beforehand as an area | region where is formed. In addition, the characteristics of the reading unit 18 may include deterioration characteristics over time in the constituent members of the reading unit 18. For example, among the constituent members of the reading unit 18, a constituent member that is expected to deteriorate with time and that has a characteristic of deterioration with time of the constituent member that is expected to affect the reading result of the image in advance is considered to be “streaked”. You may specify the area | region assumed beforehand as an area | region where a pattern is formed.

  The specific direction image extraction unit 32 has a streak defect having a streak pattern in a specific direction from the streak region cut out by the streak region cutout unit 30 (in the first embodiment, an image showing a streak defect as an example). The streak defect image) is extracted, and streak defect image information indicating the streak defect image is generated. That is, the specific direction image extraction unit 32 extracts a streak defect image having a streak pattern in a specific direction by excluding a streak pattern image in a direction other than the specific direction and a granular pattern image by halftone or the like. To generate streak defect image information. Here, HDR-VDP-2 is mentioned as another example of the extraction algorithm which excludes a granular component and extracts the streak defect of a specific direction. Other examples include an extraction algorithm using a plurality of bandpass filters, an extraction algorithm using a line detection filter, and an extraction algorithm using a differential filter (double differential filter). Hereinafter, for convenience of explanation, a streak defect image having a streak pattern in a specific direction is simply referred to as a “streaky defect”.

  FIG. 4 shows an example of the main functions of the evaluation unit 22. As shown in FIG. 4, the evaluation unit 22 includes a first image synthesis unit 34, a conversion unit 36, a specific component decomposition unit 38, a second image synthesis unit 40, and a visual model evaluation execution unit 42. The first image composition unit 34 includes a sample image indicated by the sample image information (for example, a sample image indicated by the sample image information used in the subtraction unit 28) and the streak defect image information generated by the specific direction image extraction unit 32. And the streak defect image indicated by In the first embodiment, the first image composition unit 34 synthesizes and combines the sample image and the streak defect image indicated by the streak defect image information generated by the specific direction image extraction unit 32, for example. Evaluation target image information indicating the evaluation target image which is the obtained composite image is generated. Examples of the sample image synthesized here include a sample image indicated by sample image information in which the influence of the characteristic of the reading unit 18 (for example, the characteristic of the reading execution unit 23) is excluded by the exclusion unit 24.

  Note that when the sample image and the inspection target image have the same vertical and horizontal resolutions and the specific direction resolution and the main scanning direction resolution do not match, the first image composition unit 34 and the conversion unit 36 as shown in FIG. A resolution adjusting unit 26A is interposed between the two. The resolution adjustment unit 26A aligns the vertical and horizontal resolutions of the evaluation target image indicated by the evaluation target image information and the vertical and horizontal resolutions of the sample image indicated by the sample image information. For example, for each of the evaluation target image and the sample image, the vertical and horizontal resolutions of the evaluation target image and the vertical and horizontal resolutions of the sample image are made uniform by a predetermined resolution and a resolution in the main scanning direction.

  The conversion unit 36 performs color conversion and visual filter processing on the evaluation target image indicated by the evaluation target image information and the sample image indicated by the sample image information. When the resolution adjustment unit 26A is provided, color conversion and visual filter processing are performed on the evaluation target image and the sample image whose vertical and horizontal resolutions are adjusted. Here, “color conversion” refers to conversion from a device-dependent color space to a device-independent color space. The device-dependent color space refers to, for example, a color space that depends on the printing apparatus 12 (for example, a color space that depends on spectral characteristics of a color material used for printing). The device-independent color space refers to, for example, a color space that depends on human vision (for example, a color space that depends on the spectral characteristics of the human eye), specifically, XYZ, Lab, Luv, HSV, VD. , RGB, CIECAM and the like.

  The above-mentioned “visual filter processing” means visual conversion using a function that functions as a low-pass filter for obtaining a spatial frequency corresponding to human visual characteristics (a spatial frequency predetermined as a spatial frequency close to human visual characteristics). Refers to the process of performing The “spatial frequency predetermined as a spatial frequency close to human visual characteristics” refers to a spatial frequency in the visible region excluding, for example, a spatial frequency in which stimulation to the human eye is generally strong. Here, as an example of a function that functions as a low-pass filter, a VTF function determined based on a spatial frequency characteristic corresponding to a human visual characteristic is adopted. Therefore, the “visual filter processing” here means that, for example, the evaluation target image indicated by the evaluation target image information and the sample image indicated by the sample image information do not correspond to human visual characteristics using a VTF function. This refers to processing for conversion to an evaluation target image and sample image from which the spatial frequency has been removed.

  The specific component decomposition unit 28 decomposes each of the evaluation target image and the sample image obtained by the conversion by the conversion unit 36 into specific physical quantity components. In the first embodiment, the specific component decomposition unit 28 decomposes each of the evaluation target image and the sample image into each of a plurality of spatial frequencies and each of a plurality of directions, so that the frequency for each of the plurality of spatial frequencies. A decomposition image and a direction decomposition image for each of a plurality of directions are obtained.

  The visual model evaluation execution unit 42 evaluates the visibility using the mask effect visual model for the streak defect specified by the streak defect image extracted by the extraction unit 20, and visually recognizes the streak defect. Information indicating the strength of the sex (hereinafter referred to as visibility strength information) is generated. That is, visibility is evaluated (mask effect is evaluated) for each of the frequency-resolved images and each of the direction-resolved images, and visibility intensity information is generated for each image. “Visibility strength” as used herein is represented by, for example, 10 levels of strength from 0 to 9, and the greater the value, the stronger the visibility strength (the better the visibility). Hereinafter, when it is not necessary to distinguish between the frequency-resolved image and the direction-resolved image, they are referred to as “decomposed images”.

  In the first embodiment, the visual model evaluation execution unit 42 evaluates the visibility of an image that takes into account the influence of adjacent frequency-resolved images and the influence of directional-resolved images in all directions, for example, for each decomposed image, Generate sex strength information. Here, “the influence of adjacent frequency-resolved images” refers to the influence of one frequency-resolved image on the other frequency-resolved image in a frequency-resolved image having a spatial frequency adjacent among a plurality of frequency-resolved images. Further, the “influence of the direction-resolved image in all directions” referred to here refers to the influence of a specific direction-resolved image on the other direction-resolved image in the direction-resolved image in all directions. In the first embodiment, the visual model evaluation execution unit 42 employs evaluation using VDP (visible difference predictor) as an example of an evaluation algorithm using a mask effect visual model. Here, “VDP” refers to a method of modeling human visual characteristics (masking), for example, and estimating human visible errors between two images.

  The second image synthesis unit 40 synthesizes the visibility intensity information generated for each decomposed image by the visual model evaluation execution unit 42 into the evaluation target image indicated by the evaluation target image information generated by the first image synthesis unit 34. Then, evaluation result information indicating the image obtained by the synthesis is generated. For example, the visibility intensity information of the corresponding pixel in each frequency resolved image and each direction resolved image is associated with each pixel of the evaluation target image indicated by the evaluation target image information generated by the first image composition unit 34. Here, in the frequency-resolved image, several pixels (for example, 2 × 2 pixels) of the evaluation target image appear as one pixel depending on the spatial frequency. In this case, one pixel of the frequency-resolved image corresponds to the evaluation target image. Will be associated with several pixels. Hereinafter, the image synthesized by the second image synthesis unit 40 is referred to as a “final synthesized image”.

  FIG. 5 shows an example of the main function of the warning unit 24. As shown in FIG. 5, the warning unit 24 includes a warning necessity determination unit 44 and a warning execution unit 48, and a reference information database (DB) 46 is connected to the warning necessity determination unit 44. The reference information DB 46 stores information (hereinafter referred to as “reference information”) that is referred to when determining whether or not a warning is necessary. The warning necessity determination unit 44 refers to the reference information in the reference information DB 46 and determines whether or not a warning is necessary based on the evaluation result information generated by the visual model evaluation execution unit 42. In the reference information DB 46 according to the first embodiment, determination information indicating whether or not visual recognition is possible is stored for each visibility strength as an example of reference information. Therefore, the warning necessity determination unit 44 can determine, based on the determination information, whether or not the visibility strength indicated by the visibility strength information included in the evaluation result information is a strength that requires a warning. Note that the “determination information” referred to here is determined in advance based on, for example, a subjective evaluation result by a specific person (for example, an evaluation result by a sensory test). It may be determined in advance based on the result.

  When the warning necessity determining unit 44 determines that the visibility intensity indicated by the visibility intensity information included in the evaluation result information requires a warning (the streak identified by the streak defect image) Warning if the defect is evaluated as a visible streak defect). In the first embodiment, as a warning mode by the warning execution unit 48, for example, a warning mode is provided by directly processing the inspection target printed matter 16 including the streak defect evaluated as a visible streak defect. And a mode in which predetermined information is visibly displayed. An example of “direct processing” here is sorting processing. In addition to this, for example, a process for providing a specific support mark for supporting the specification of the position of a streak defect in the inspection object print 16 or a specific support mark at a position where the streak defect in the inspection object print 16 is extended in a specific direction. The process which provides can be illustrated.

  The extraction unit 20, the evaluation unit 22, and the warning unit 24 are realized by, for example, the computer 60 and other input / output devices shown in FIG. FIG. 6 shows an example of the main configuration of the electrical system of the image evaluation apparatus 10. Hereinafter, when it is not necessary to distinguish between the sample printed matter 14 and the inspection target printed matter 16, they are simply referred to as “printed matter”.

  As shown in FIG. 6, the computer 60 includes a CPU (Central Processing Unit) 62, a memory 64, and a nonvolatile storage unit 66, which are connected via a bus 68 including an address bus, a system bus, and the like. Are connected to each other. The storage unit 66 is realized by an SSD (Solid State Drive), an HDD (Hard Disk Drive), or the like. An image evaluation processing program 70 is stored in the storage unit 66 as a storage medium. The storage unit 66 includes a first storage area 72, a second storage area 74, and a reference information DB 46. Sample image information is stored in the first storage area 72, and inspection target image information is stored in the second storage area 74.

  The CPU 62 reads out the image evaluation processing program 70 from the storage unit 66 and develops it in the memory 64, and sequentially executes processes included in the image evaluation processing program 70. The image evaluation processing program 70 has an extraction process 76, an evaluation process 78, and a warning process 80. The CPU 62 operates as the extraction unit 20 illustrated in FIG. 1 by executing the extraction process 76. The CPU 62 operates as the evaluation unit 22 illustrated in FIG. 1 by executing the evaluation process 78. The CPU 62 operates as the warning unit 24 illustrated in FIG. 1 by executing the warning process 80.

  Although the case where the image evaluation processing program 70 is read from the storage unit 66 is illustrated here, it is not necessarily stored in the storage unit 66 from the beginning. For example, the image evaluation processing program 70 is first stored in an arbitrary “portable storage medium” such as a flash memory, a CD-ROM, a DVD disk, a magneto-optical disk, and an IC card that is connected to the computer 60. It may be left. Then, the computer 60 may acquire and execute each program from these portable storage media. Further, the image evaluation processing program 70 may be stored in an external device such as another computer or a server device connected to the computer 60 via the Internet or a LAN (Local Area Network). In this case, the computer 60 may acquire and execute the image evaluation processing program 70 from the external device.

  The image evaluation apparatus 10 includes an input / output device. Further, the image evaluation apparatus 10 is an input / output interface (I) that electrically connects the computer 60 and various input / output devices to control transmission and reception of various information between the computer 60 and various input / output devices. / O) 82. The input / output device is electrically connected to the computer 30 via the bus 68 by being connected to the I / O 82. Here, as an input / output device, a receiving unit 84, a display unit 86, a recording unit 88, a communication interface (I / F) 90, a transport unit 92, a scanner 94 as an example of the reading unit 18, and a sorter 96 are applied. Yes.

  The accepting unit 84 accepts an operation input by a user of the image evaluation device 10. Examples of the receiving unit 84 include input devices such as a keyboard, a mouse, a transparent touch panel that is used over the display, an operation button for turning on the power, an operation button for setting various information, and a scroll key.

  The display unit 86 displays various information. An example of the display unit 86 is a liquid crystal display. In the first embodiment, a touch panel display formed by superimposing a touch panel which is a part of the receiving unit 84 on a liquid crystal display as the display unit 86 is employed.

  The recording unit 88 records an image on the inspection target printed material 16. In the first embodiment, an ink jet recording head is employed as an example of the recording unit 88, but the disclosed technology is not limited thereto. For example, a thermal printer or a stamp device may be used, and any device can be applied as long as it can apply a specific mark to the inspection target printed matter 16.

  The communication I / F 90 is connected to an external device 93 such as a personal computer or a printer via a communication network 91 and controls transmission / reception of various information between the external device 93 and the computer 60.

  The transport unit 92 takes the printed material into the image evaluation apparatus 10 and transports the captured printed material along a predetermined transport path, thereby passing the reading position of the scanner 94 and the recording position of the recording unit 88 to the sorter 96. Send it in. The conveyance unit 92 includes a motor 98 that is used as a drive source for conveying the printed material. The transport unit 92 includes a driver 100 that is connected to the motor 98 and controls the driving of the motor 98. The driver 100 is connected to the I / O 82. Therefore, the computer 60 can control the driving of the motor 98 via the driver 100.

  The scanner 94 optically reads an image included in the printed material and outputs image information indicating the read image to the computer 60. In the first embodiment, a reduction optical system type scanner is employed as an example of the scanner 94. However, the present invention is not limited to this. For example, an equal magnification optical system type scanner may be employed.

  FIG. 7 shows an example of the main configuration of the electrical system of the scanner 94. As shown in FIG. 7, the scanner 94 includes a computer 94 </ b> A that is an example of the exclusion unit 25. The computer 94A includes a CPU 94B, a memory 94C, and a non-volatile storage unit 94D, which are connected to each other via a bus 94E that includes an address bus, a system bus, and the like. Note that the storage unit 94D is realized by an SSD, an HDD, or the like. A sample image reading processing program 95A and an inspection object image processing program 95B are stored in the storage unit 94D as a storage medium.

  The CPU 94B reads the sample image reading processing program 95A from the storage unit 94D, develops it in the memory 94C, and sequentially executes the processes included in the sample image reading processing program 95A. Further, the CPU 94B reads out the inspection object image reading processing program 95B from the storage unit 94D, develops it in the memory 94C, and sequentially executes the processes of the inspection object image reading processing program 95B.

  The scanner 94 includes an input / output device. The scanner 94 includes an I / O 94F that electrically connects the computer 94A and various input / output devices to control transmission / reception of various information between the computer 94A and the various input / output devices. The input / output device is electrically connected to the computer 94A via the bus 94E by being connected to the I / O 94F. Here, an image sensor 94F, a light irradiation unit 94G, and an external I / F 94H, which are examples of the reading execution unit 23, are applied as input / output devices. The light irradiation unit 94G irradiates light onto a reading target surface (a surface on which a reading target image is printed) of the printed matter. In the first embodiment, a white fluorescent lamp is applied as an example of the light irradiation unit 94G, but another light source may be used.

  The image sensor 94F receives the light reflected on the reading target surface of the printed matter by the light irradiation unit 94G and receives the light reflected on the reading target surface, and outputs image information obtained by photoelectric conversion to the computer 94A. The computer 94A stores the input image information in the storage unit 94D. In the first embodiment, a CCD (Charge Coupled Devices) image sensor is applied as an example of the image sensor 94F, but another image sensor may be used.

  The external I / F 94H is connected to the I / O 82 shown in FIG. 6 and controls transmission / reception of various information between the computer 94A and the computer 60.

  The sorter 96 shown in FIG. 6 performs a sorting process. That is, the printed material conveyed by the conveyed product 92 is taken in and sorted according to the type of the printed material. For example, the sample printed matter 14 and the inspection subject printed matter 16 are sorted. Further, the captured printed material 16 to be inspected is sorted based on the visibility of the inspection target image. In the first embodiment, two types of sorting methods are adopted as the sorting method of the printed material 16 to be inspected, and these sorting methods are properly used according to a user instruction. FIG. 8 schematically shows an example of two types of sorting methods. As shown in FIG. 8, one of the two sort methods is a “shifting method” and the other is a “tray-specific method”. In the “shifting method”, the inspection target printed matter 16 that is evaluated as having good visibility is accumulated in a predetermined alignment direction (for example, the vertical direction) in the discharge destination tray of the inspection target printed matter 16, and the visibility is evaluated as poor. This means a sorting method in which the printed material 16 to be inspected is removed from a predetermined alignment direction. The “tray-specific method” is a method in which the inspection target printed matter 16 with good visibility of the inspection target image and the inspection target printed matter 16 with poor visibility of the inspection target image are discharged to different areas (for example, a tray). Point to.

  FIG. 9 schematically shows an arrangement example of the recording unit 88, the scanner 94, and the sorter 96. As shown in FIG. 9, a plurality of conveyance roll pairs 102 are arranged on the conveyance path of the printed material. In addition, a scanner 94 and a recording unit 88 are sequentially arranged on the conveyance path from the upstream side in the conveyance direction of the printed material, and a sorter 96 is disposed at the end of the conveyance route so as to receive the printed material. The pair of transport rolls 102 includes transport rolls 102A and 102B that rotate in directions opposite to each other under the rotational driving force of the motor 98. The conveyance roll pair 102 sandwiches the printed material between the conveyance roll 102 </ b> A and the conveyance roll 102 </ b> B, conveys the printed material to the reading position by the scanner 94 and the recording position by the recording unit 88, and sends the printed material to the sorter 96. In the first embodiment, the orientation of the printed material during conveyance in the image evaluation device 10 is the same as the direction of the printed material during conveyance in the printing device 12.

  As shown in FIGS. 10 to 12, the recording unit 88 is an example of the case where the inspection target printed matter 16 that is evaluated as having poor visibility is conveyed along the conveyance direction corresponding to the specific direction according to the disclosed technology. The specific support mark for specifying the position of the streak defect is recorded with respect to the inspection object printed matter 16. In the example illustrated in FIG. 10, a blank area (a downstream side of the outer peripheral edge in the transport direction) that is a position where the streak defect extends in the transport direction (here, as an example, downstream in the transport direction) and downstream of the inspection target print 16 in the transport direction. The specific support mark is recorded at the position of a predetermined margin area). In the example shown in FIG. 11, a blank area (a predetermined outer peripheral edge is a position where the streak defect extends in the conveyance direction (here, as an example, upstream in the conveyance direction) and upstream in the conveyance direction of the print 16 to be inspected. A specific support mark is recorded at the position of the margin area. In the example shown in FIG. 12, the specific support mark is recorded at the same position as in the example shown in FIG. 10, and further, the printed matter to be inspected is a position where the streak defect extends in the direction intersecting the transport direction. A specific support mark is recorded at the position of 16 blank areas. Note that the “direction intersecting the transport direction” referred to herein indicates a direction orthogonal to the transport direction, for example.

  In the examples shown in FIGS. 10 to 12, the red rectangular mark is applied as the specific support mark. However, the present invention is not limited to this, and marks of other shapes with other colors (preferably different colors from the recording paper) are used. May be adopted.

  Moreover, when the printed material 16 to be inspected includes a plurality of streak patterns having different visibility intensities (for example, area and density), a mark of a different color may be given to each streak pattern. For example, as shown in FIG. 13, when the inspection target printed matter 16 includes two streak patterns having different visibility intensities, each of the streaky patterns extends in the transport direction and is the position of the inspection target printed matter 16. A red mark and a blue mark are selectively recorded at the position of the blank area according to the intensity. In addition, the red mark and the blue mark are intensified at the position of the blank area of the printed material 16 to be inspected, which is a position where each of the streaks is extended in a direction intersecting the conveyance direction (in this example, a direction orthogonal thereto). Selectively record according to. In the example shown in FIG. 13, the red mark has higher visibility than the blue mark.

  In the first embodiment, an example in which marks are recorded by an ink jet recording head has been described. However, the disclosed technique is not limited thereto. For example, a stamp or a seal may be employed as the specific support mark. If a stamp is used, a stamping device capable of stamping may be installed instead of the recording unit 88. If a seal is used, a seal sticking device capable of sticking a seal may be installed instead of the recording unit 88. Good.

  Next, as an operation of the first embodiment, image evaluation processing performed by the image evaluation apparatus 10 when the CPU 62 executes the image evaluation processing program 70 will be described with reference to FIG. Here, in order to avoid complications, the sample printed matter 14 and the inspection subject printed matter 16 are different from each other in waiting areas (for example, different supply areas) so as to be taken into the image evaluation apparatus 10 one by one in accordance with an instruction from the user. The case where it is already stored in the paper tray will be described. Further, here, in order to avoid complications, an example in which all of the sample printed matter 14 printed by the printing apparatus 12 is accommodated in the standby area and all of the sample printed matter 14 accommodated in the standby area is read by the scanner 94 is described as an example. Will be described. Further, here, in order to avoid complications, the description will be made on the assumption that the orientation of the printed material taken in and conveyed by the image evaluation device 10 is the same as the orientation of the printed material on the conveying path in the printing device 12. . In addition, here, in order to avoid complications, a case will be described in which the “shift method” or “tray-specific method” has already been instructed by the user as the sorting method.

  In the image evaluation process shown in FIG. 14, first, in step 200, the reading control process is performed by the extraction unit 20, and then the process proceeds to step 202. FIG. 15 shows an example of the flow of reading control processing. In the reading control process shown in FIG. 15, first, in step 200 </ b> A, it is determined whether or not the extracting unit 20 has received an instruction to start reading a sample image by the receiving unit 84. When the instruction to start reading the sample image is accepted by the accepting unit 84 in step 200A, the determination is affirmed and the process proceeds to step 200B. If the instruction to start reading the sample image is not accepted by the accepting unit 84 in step 200A, the determination is denied and the determination in step 200A is performed again.

  In step 200 </ b> B, the extraction unit 20 instructs the conveyance unit 92 to take in one sample printed material 14. In response to this, the conveyance unit 92 takes in one sample printed matter 14 and conveys it along the conveyance path.

  In the next step 200 </ b> C, the extraction unit 20 determines whether the sample printed matter 14 has reached the reading position by the scanner 94. If the sample printed matter 14 has reached the reading position by the scanner 94 in step 200C, the determination is affirmed and the routine proceeds to step 200D. If the sample printed matter 14 has not reached the reading position by the scanner 94 in step 200C, the determination is negative and the determination in step 200C is performed again.

  In step 200D, the extraction unit 20 instructs the scanner 94 to start reading the sample image included in the sample printed matter 14, and then the process proceeds to step 200E. In step 200E, the extraction unit 20 determines whether the reading of the sample image by the scanner 94 is completed. If the reading of the sample image by the scanner 94 is completed in step 200E, the determination is affirmed and the process proceeds to step 200F. If the reading of the sample image by the scanner 94 is not completed in step 200E, the determination is denied and the determination in step 200E is performed again.

  In step 200F, the extraction unit 20 determines whether or not the processing in steps 200B to 200E has been performed for all the sample printed matter 14 accommodated in the standby area. If the processing of Step 200B to Step 200E is not performed for all the sample printed materials 14 in Step 200F, the determination is negative and the processing proceeds to Step 200B. In step 200F, when the processing of step 200B to step 200E is performed for all the sample printed matter 14, the determination is affirmed and the process proceeds to step 200G.

  In step 200G, the extraction unit 20 determines whether an instruction to start reading the inspection target image has been received by the reception unit 84. If the reception unit 84 receives an instruction to start reading the inspection target image in step 200G, the determination is affirmed and the process proceeds to step 200H. If the instruction to start reading the inspection target image is not received by the receiving unit 84 in step 200G, the determination is denied and the determination in step 200G is performed again.

  In step 200 </ b> H, the extraction unit 20 instructs the conveyance unit 92 to take in one printed material 16 to be inspected. In response to this, the conveyance unit 92 takes in one printed material 16 to be inspected and conveys it along the conveyance path.

  In the next step 200I, the extraction unit 20 determines whether or not the inspection object printed matter 16 has reached the reading position by the scanner 94. If it is determined in step 200I that the printed material 16 to be inspected has reached the reading position by the scanner 94, the determination is affirmed and the process proceeds to step 200J. If it is determined in step 200I that the printed material 16 to be inspected has not reached the reading position by the scanner 94, the determination is negative and the determination in step 200I is performed again.

  In step 200J, the extraction unit 20 instructs the scanner 94 to start reading the inspection target image included in the inspection target printed matter 16, and then proceeds to step 200K. In step 200K, the extraction unit 20 determines whether reading of the inspection target image by the scanner 94 is completed. If the reading of the inspection target image by the scanner 94 is completed in step 200K, the determination is affirmed and the process proceeds to step 200L. If the reading of the inspection target image by the scanner 94 is not completed in step 200K, the determination is negative and the determination in step 200K is performed again.

  In step 200L, the extraction unit 20 determines whether or not the processing in steps 200H to 200K has been performed for all the inspection target prints 16 accommodated in the standby area. If the processing of Step 200H to Step 200K is not performed for all the inspection target printed materials 16 in Step 200L, the determination is negative and the processing proceeds to Step 200H. If the processes of Step 200H to Step 200K are performed for all the inspection target printed materials 16 in Step 200L, the determination is affirmed and the reading control process ends.

  Next, a sample image reading process performed by the scanner 94 when the CPU 94B of the scanner 94 executes the sample image reading processing program 95A will be described with reference to FIG.

  In the sample image reading process shown in FIG. 16, first, in step 210, the exclusion unit 25 determines whether or not the start of reading of the sample image is instructed in step 200D. If the start of reading the sample image is not instructed in this step 210, the determination is denied and the determination in this step 210 is performed again. If the start of reading of the sample image is instructed in step 210, the determination is affirmed and the routine proceeds to step 212.

  In step 212, the exclusion unit 25 activates the image sensor 94F to cause the image sensor 94F to start reading a sample image. As a result, the image sensor 94F acquires sample image information and outputs the acquired sample image information to the computer 94A.

  In the next step 214, the exclusion unit 25 eliminates the influence of the characteristics of the scanner 94 from the sample image information input from the image sensor 94F. In step 214, for example, the exclusion unit 25 excludes the influence of the characteristics of the image sensor 94F from the sample image information input from the image sensor 94F.

  In the next step 216, the sample image information from which the influence of the characteristics of the image sensor 94F has been eliminated in step 214 is stored in the storage unit 94D by the exclusion unit 25, and then the process proceeds to step 218. In step 218, the exclusion unit 25 determines whether the reading of the sample image has been completed. If the reading of the sample image is not completed in step 218, the determination is negative and the routine proceeds to step 212. If the reading of the sample image is completed in step 218, the determination is affirmed and the routine proceeds to step 220.

  In step 220, the sample image information stored in the storage unit 94 </ b> D in step 216 is output to the computer 60 of the image evaluation apparatus 10 by the exclusion unit 25.

  In the next step 222, the exclusion unit 25 determines whether or not the processing in steps 210 to 220 has been performed on all the sample printed matter 14 accommodated in the standby area. If it is determined in step 222 that the processing in steps 210 to 220 has not been performed for all the sample printed materials 14, the determination is negative and the processing proceeds to step 210. When the processing of Step 210 to Step 220 is performed for all the sample printed materials 14 in Step 222, the determination is affirmed and the sample image reading processing is terminated.

  Next, an inspection target image reading process performed by the scanner 94 when the CPU 94B of the scanner 94 executes the inspection target image reading processing program 95B will be described with reference to FIG.

  In the inspection object image reading process shown in FIG. 17, first, in step 230, the exclusion unit 25 determines whether or not the start of the inspection object image reading is instructed in step 200J. If the start of reading of the inspection target image is not instructed in step 230, the determination is negative and the determination in step 230 is performed again. If it is instructed to start reading the inspection target image in step 230, the determination is affirmed and the process proceeds to step 232.

  In step 232, the exclusion unit 25 operates the image sensor 94F to cause the image sensor 94F to start reading the inspection target image. As a result, the image sensor 94F acquires the inspection target image information and outputs the acquired inspection target image information to the computer 94A.

  In the next step 234, the exclusion unit 25 eliminates the influence of the characteristics of the scanner 94 from the inspection target image information input from the image sensor 94F. In step 234, for example, the exclusion unit 25 excludes the influence of the characteristics of the image sensor 94F from the inspection target image information input from the image sensor 94F.

  In the next step 236, the inspection unit image information in which the influence of the characteristics of the image sensor 94F in step 234 is excluded by the exclusion unit 25 is stored in the storage unit 94D, and then the process proceeds to step 238. In step 238, the exclusion unit 25 determines whether reading of the inspection target image has been completed. If the reading of the inspection target image has not been completed in step 238, the determination is negative and the process proceeds to step 232. If the reading of the inspection target image is completed in step 238, the determination is affirmed and the routine proceeds to step 240.

  In step 240, the exclusion unit 25 outputs the inspection target image information stored in the storage unit 94D in step 236 to the computer 60 of the image evaluation apparatus 10.

  In the next step 242, the exclusion unit 25 determines whether or not the processing in steps 230 to 240 has been performed for all the inspection target printed materials 16 accommodated in the standby area. If it is determined in step 242 that the processing in steps 230 to 240 has not been performed for all the inspection target printed materials 16, the determination is negative and the processing proceeds to step 230. If the processing of Step 230 to Step 240 is performed for all the inspection target printed matter 16 in Step 242, the determination is affirmed and the inspection target image reading processing is ended.

  Returning to FIG. 14, in the image evaluation process, the extraction unit 20 performs the extraction process in step 202, and then proceeds to step 204. FIG. 18 shows an example of the flow of extraction processing. In the extraction process shown in FIG. 18, first, in step 202A, the extraction unit 20 determines whether image information is input from the scanner 94 or not. If no image information is input from the scanner 94 in step 202A, the determination is denied and the determination in step 202A is performed again. If image information is input from the scanner 94 in step 202A, the determination is affirmed and the process proceeds to step 202B.

  In step 202B, the extraction unit 20 determines whether the image information input from the scanner 94 in step 202A is sample image information. If the image information input from the scanner 94 is sample image information in this step 202B, the determination is affirmed and the routine proceeds to step 202C. If the image information input from the scanner 94 is not sample image information in this step 202B (if the image information input from the scanner 94 is image information to be inspected), the determination is negative and the routine proceeds to step 202D.

  In step 202C, the sample image information input from the scanner 94 by performing step 220 is stored in the first storage area 72 by the subtracting unit 28, and then the process proceeds to step 202F.

  In step 202D, the image information to be inspected input from the scanner 94 is stored in the second storage area 74 by the resolution adjustment unit 26 by performing the above step 240, and then the process proceeds to step 202E. In step 202E, the resolution adjustment unit 26 adjusts the resolution of the inspection target image indicated by the inspection target image information stored in the second storage area 74, and then the process proceeds to step 202F. In this step 202E, for example, the resolution adjustment unit 26 performs interpolation processing on the pixels in the characteristic direction of the inspection target image, thereby aligning the resolution in the specific direction of the inspection target image with the resolution in the main scanning direction.

  In step 202F, the subtraction unit 28 calculates a difference value between the sample image information and the inspection target image information. In this step 202F, for example, the subtraction unit 28 uses the difference value between the sample image indicated by the sample image information stored in the first storage area 72 in step 202C and the inspection target image whose resolution has been adjusted in step 202E. Is calculated. The “difference value” here refers to, for example, the absolute value of the difference between the density of the sample image and the density of the inspection target image. Note that the difference value is not limited to the density difference value, and may be a reflectance difference value. Further, it may be at least one of a density difference, a color space difference, a light amount difference, a luminance difference, and a device signal value difference between the inspection target image and the virtual sample image.

  In the next step 202G, the streak region cutout unit 30 cuts out a streak region targeting a predetermined region of the inspection target image whose resolution has been adjusted in the above step 202E. Move to 202H. In step 202H, the specific direction image extraction unit 32 has a streak pattern in which the difference value calculated in step 202F exceeds a predetermined value in the streak region cut out in step 202G, and a streak pattern in a specific direction is present. It is determined whether or not it exists. Here, the “predetermined value” refers to a value corresponding to a difference value that is predetermined as a difference value of a streak pattern recognized as an image defect, for example, and is a value obtained by an experiment or simulation using an actual machine. . If the streak pattern in the step 202H is a streak pattern in which the difference value calculated in the above step 202F exceeds a predetermined value and there is no streak pattern in a specific direction, the determination is negative and shown in FIG. The image evaluation process ends. In step 202H, if the streak pattern is a streak pattern in which the difference value calculated in step 202F exceeds a predetermined value and a streak pattern in a specific direction exists, the determination is affirmed and the process proceeds to step 202I. To do.

  In step 202I, the specific direction image extraction unit 32 extracts a streak defect image in a specific direction, and thereafter ends the extraction process.

  Returning to FIG. 14, in the image evaluation process, the evaluation unit 22 performs the evaluation process in step 204, and then the process proceeds to step 206. FIG. 19 shows an example of the flow of evaluation processing. In the evaluation process shown in FIG. 19, first, in step 202J, the first image composition unit 34 acquires sample image information from the first storage area 72, and then the process proceeds to step 202K. In step 202K, the first image synthesis unit 34 synthesizes the streak defect image extracted in step 202I and the sample image indicated by the sample image information acquired in step 202J, thereby evaluating image information to be evaluated. Is generated. In the next step 202L, the first image composition unit 34 stores the evaluation target image information generated in step 202K in the storage unit 66.

  In the next step 204A, the evaluation unit image information is acquired from the storage unit 66 by the conversion unit 36, and then the process proceeds to step 204B. In step 204B, the conversion unit 36 performs color conversion and color conversion on the evaluation target image indicated by the evaluation target image information acquired in step 204A and the sample image indicated by the sample image information acquired in step 202J. Visual filter processing is performed.

  In the next step 204C, each of the evaluation target image and the sample image obtained by performing the color conversion and the visual filter processing in the above step 204B by the specific component decomposing unit 38 is converted into each of a plurality of spatial frequencies and a plurality of directions. Each is broken down. Thereby, a plurality of frequency-resolved images and a plurality of direction-resolved images are generated.

  In the next step 204D, the visual model evaluation execution unit 42 evaluates the visibility using the mask effect visual model for the streak defect specified by the streak defect image extracted in step 202I. Visibility intensity information is generated. Specifically, the visibility intensity information is evaluated by evaluating the visibility of the image in which the influence of the adjacent frequency-resolved image and the influence of the directional-resolved image in all directions are added to each decomposed image obtained in step 204C. Is generated.

  In the next step 204E, the second image synthesis unit 40 synthesizes the plurality of frequency-decomposed images and the plurality of direction-decomposed images with the evaluation target image indicated by the evaluation target image information acquired in step 204A. Evaluation result information indicating the final composite image is generated.

  In the next step 204F, the visual model evaluation execution unit 42 stores the evaluation result information generated in step 204E in the storage unit 66, and then ends the evaluation process.

  Returning to FIG. 14, in the image evaluation processing, warning processing is performed by the warning unit 24 in step 206, and then the image evaluation processing is terminated. FIG. 20 shows an example of the flow of warning processing. In the warning process shown in FIG. 20, first, in step 206A, the warning necessity determination unit 44 acquires the evaluation result information from the storage unit 55, and then the process proceeds to step 206B. In step 206B, the warning necessity determination unit 44 acquires reference information from the reference information DB 46, and then the process proceeds to step 206C.

  In step 206C, the warning necessity determination unit 44 refers to the reference information in the reference information DB 46, and the strength of the visibility of the streak defect specified by the streak defect image extracted in step 202I is a strength that requires a warning. It is determined whether or not. In this step 206, for example, the warning necessity determination unit 44 refers to the determination information that is an example of the reference information, and the strength of the visibility indicated by the visibility strength information included in the evaluation result information is a strength that requires a warning. It is determined whether or not. If it is determined in step 206 that the intensity does not require a warning, the determination is denied and the warning process is terminated. If it is the intensity that requires a warning in step 206, the determination is affirmed and the routine proceeds to step 206D.

  In step 206D, the warning is executed by the warning execution unit 48. Here, for example, when the “shifting method” is instructed, the warning execution unit 48, as an example, displays the inspection target print 16 having an unacceptable streak pattern as shown in FIG. The sorter 96 is controlled so as to be removed from (for example, ejected) 16. Further, for example, when “tray-specific method” is instructed, the warning execution unit 48 separates the inspection target printed matter 16 having an unacceptable streak pattern and the other inspection target printed matter 16 as shown in FIG. The sorter 96 is controlled so that it can be sorted.

  In Step 206D, the warning execution unit 48 evaluates that the streak defect specified by the streak defect image is a visible streak defect (by the visibility intensity information included in the evaluation result information). When the visibility intensity shown is equal to or higher than the predetermined intensity), the predetermined information is displayed on the display unit 86. Here, the “predetermined information” refers to specific information that identifies the inspection target printed matter 16 including the streak defect evaluated as a streak defect that can be visually recognized by the evaluation unit 22, for example.

  As an example of the specific information, for example, a defective page number shown in FIG. The “defective page number” referred to here is a serial number of the inspection object printed matter 16 from which the inspection object image is read by the scanner 94. That is, the “defective page number” refers to the page number of the inspection target printed matter 16 that includes the streak defect that is evaluated as a streak defect that can be visually recognized by the evaluation unit 22. . In the example illustrated in FIG. 21, the visibility strength of the streak defect included in the inspection target printed matter 16 specified by the defective page number is displayed on the display unit 86 together with the defective page number. In the example shown in FIG. 21, the inspection target image included in the inspection target printed matter 16 specified by the defective page number is displayed on the display unit 86, and a numerical value indicating the degree of the defect (the visibility of the streak defect). A numerical value corresponding to the intensity) is displayed on the display unit 86 together with the inspection target image. In the example illustrated in FIG. 21, a specific support image that supports the specification of the position of the streak defect in the inspection target image displayed on the display unit 86 is superimposed on the inspection target image and displayed on the display unit 86. . The specific support image refers to, for example, a red mark and a mark (a mark corresponding to the degree of defect) corresponding to the red mark and the blue mark shown in FIG. 13, and in this case, each mark is a position of a streak defect on the inspection target image. Is displayed at a position where it can be specified.

  In the first embodiment, the degree of the streak defect can be specified according to the color of the mark. However, the present invention is not limited to this, and the degree of the streak defect is specified by highlight display or the size of the mark. You may be able to do it. In addition, a mode in which the level of the streak defect is lightly displayed and the case of a severe streak is displayed blinking can be exemplified.

  In the first embodiment, the visual display by the display unit 86 is exemplified, but the disclosed technique is not limited to this, and may be a permanent display by a printer or an audible display by an audio reproduction device, for example. . Moreover, the display which combined at least 2 of the visible display by the display part 86, the permanent visible display by a printer, and the audible display by an audio reproduction apparatus may be sufficient. Moreover, you may perform a visual display using display parts, such as a portable terminal device and a tablet terminal device. Further, the communication means used for causing the display unit to perform visible display is not limited to wired communication, and may be wireless communication using BlueTooth (registered trademark), wireless LAN (Local Area Network), or the like.

  FIG. 38 shows an example of an evaluation result when image evaluation processing is performed on a sample image and an inspection target image that do not have a granular pattern, and image evaluation on a sample image and an inspection target image that have a granular pattern. An example of an evaluation result when processing is performed is schematically shown. As shown in FIG. 38, when visibility is evaluated using a visual model with a mask effect for an inspection target image having a granular pattern, it is expressed by a simple image difference (difference value between the sample image and the inspection target image). No visible streak pattern is detected at the position of the streak pattern. In this case, the subjective evaluation by the subject is an evaluation that the streak pattern cannot be visually recognized. On the other hand, when the visibility is evaluated using the mask effect visual model for the inspection target image that does not have a granular pattern, the level of the level that can be visually recognized at substantially the same position as the streak pattern obtained by the simple image difference. A streak pattern is detected. In this case, the subjective evaluation by the subject is an evaluation that the streak pattern can be easily visually recognized. In the example shown in FIG. 38, when the visibility is evaluated using a visual model without a mask effect for the inspection target image regardless of the presence or absence of the granular pattern, the position of the streak pattern expressed by the simple image difference A mode in which a streak pattern is detected at substantially the same position is shown.

  As described above, in the image evaluation apparatus 10 according to the first embodiment, the extraction unit 20 uses the sample image information read by the reading unit 18 and the inspection contrast image information read by the reading unit 18. Based on the difference value, a streak defect image is extracted from the inspection target image. Since the evaluation unit 22 evaluates the visibility of the streak defect specified by the streak defect image extracted by the extraction unit 20, the evaluation unit 22 has an unacceptable streak defect as compared to the case where this configuration is not provided. It can be determined with high accuracy whether the printed material 16 is the inspection object.

  Further, in the image evaluation apparatus 10 according to the first embodiment, as a streak defect specified by a streak defect image to be evaluated by the evaluation unit 22, a streak defect having a spatial frequency corresponding to human visual characteristics is used. Is adopted. Thereby, compared with the case where it does not have this structure, the evaluation accuracy of the visibility of a streak defect can be improved.

  Further, in the image evaluation apparatus 10 according to the first embodiment, the streak defect in the color space corresponding to the human visual characteristic is specified as the streak defect specified by the streak defect image to be evaluated by the evaluation unit 22. Is adopted. Thereby, compared with the case where it does not have this structure, the evaluation accuracy of the visibility of a streak defect can be improved.

  Further, in the image evaluation apparatus 10 according to the first embodiment, a streak defect is extracted with respect to a predetermined region of the inspection target image indicated by the inspection target image information. As a result, the processing load can be reduced as compared with the case without this configuration.

  Further, in the image evaluation apparatus 10 according to the first embodiment, the streak defect is extracted by targeting a region assumed in advance as a region where the streak defect is formed. Thereby, compared with the case where it does not have this structure, the fall of evaluation accuracy can be suppressed, reducing the load concerning a process.

  Further, in the image evaluation device 10 according to the first embodiment, the region that is assumed in advance as the region where the streak defect is formed is specified based on at least one characteristic of the printing device 12 and the reading unit 18. As an area. Thereby, compared with the case where it does not have this structure, the streak defect of a specific direction can be extracted with high precision with a simple structure.

  Further, in the image evaluation apparatus 10 according to the first embodiment, the printing apparatus 12 is an element of the characteristics of the printing apparatus 12 that is used to specify an area that is assumed in advance as an area in which a streak defect is formed. The characteristic of deterioration with time of the structural member is adopted. Thereby, compared with the case where it does not have this structure, the streak defect of a specific direction can be extracted still more accurately.

  Further, in the image evaluation apparatus 10 according to the first embodiment, the reading unit 18 is used as one element of the characteristics of the reading unit 18 that is used for specifying a region that is assumed in advance as a region in which a streak defect is formed. The characteristic of deterioration with time of the structural member is adopted. Thereby, compared with the case where it does not have this structure, the streak defect of a specific direction can be extracted still more accurately.

  In the image evaluation apparatus 10 according to the first embodiment, the recording paper transport direction in the printing apparatus 12 is adopted as the specific direction. Thereby, compared with the case where it does not have this structure, the streak defect of a specific direction can be extracted still more accurately.

  In the image evaluation device 10 according to the first embodiment, the printing method of the printing device 12 is a single pass method. Thereby, compared with the case where this structure is not provided, the streak defect of a specific direction can be extracted with high precision.

  In the image evaluation apparatus 10 according to the first embodiment, the inspection target image is read so that the resolution in the specific direction is lower than the resolution in the direction intersecting the specific direction. Thereby, compared with the case where this structure is not provided, the load concerning a reading process can be reduced.

  In the image evaluation apparatus 10 according to the first embodiment, after the inspection target image is read, the sample image information and the inspection control image are obtained after the resolution in the specific direction is aligned with the resolution in the direction intersecting the specific direction. A difference value from the information is calculated. Thereby, compared with the case where it does not have this structure, the streak defect of a specific direction can be extracted still more accurately.

  In the image evaluation apparatus 10 according to the first embodiment, the influence of the characteristics of the reading unit 18 is excluded from at least the inspection target image information of the sample image information and the inspection target image information, and the influence of the characteristics of the reading unit 18 is eliminated. The difference value is calculated using the inspection target image information from which the is excluded. Thereby, compared with the case where it does not have this structure, the streak defect of a specific direction can be extracted still more accurately.

  In the image evaluation apparatus 10 according to the first embodiment, the evaluation unit 22 evaluates the visibility of the streak defect using a mask effect visual model. Thereby, compared with the case where this configuration is not provided, it is possible to determine with higher accuracy whether or not the inspection target image 16 having the unacceptable streak defect is printed.

  Further, the image evaluation apparatus 10 according to the first embodiment is based on the evaluation target image obtained by combining the streak defect image with the sample image by the first image combining unit 34 and the sample image indicated by the sample image information. It has the structure which evaluates the visibility of a streak defect using a mask effect visual model. Thereby, compared with the case where it does not have this structure, the evaluation accuracy of the visibility of a streak defect can be improved.

  Further, the image evaluation apparatus 10 according to the first embodiment decomposes the sample image indicated by the evaluation target image and the sample image information with respect to each of a plurality of spatial frequencies and each of a plurality of directions, thereby obtaining a plurality of spaces. The frequency-resolved image for each frequency and the direction-resolved image for each of a plurality of directions are obtained. Thereby, compared with the case where it does not have this structure, the evaluation accuracy of the visibility of a streak defect can be improved.

  Further, the image evaluation apparatus 10 according to the first embodiment has a configuration that warns when it is determined that the streak defect whose visibility is evaluated by the evaluation unit 22 is a visible streak defect. Yes. Thereby, compared with the case where this configuration is not provided, it is possible to easily grasp that the inspection target printed matter 16 is printed with the inspection target image having an unacceptable streak defect.

  In addition, the image evaluation apparatus 10 according to the first embodiment has a configuration in which a warning is given by directly performing processing on the printed material 16 to be inspected including a streak defect that is determined to be a visible streak defect. doing. Thereby, compared with the case where this configuration is not provided, it is possible to specify the inspection target printed matter 16 on which the inspection target image having unacceptable streak defects is printed with a simple configuration.

  In addition, the image evaluation apparatus 10 according to the first embodiment has a configuration in which a warning is given by giving a specific support mark to the inspection target printed matter 16 to support specification of the position of the streak defect in the inspection target printed matter 16. Yes. Thereby, compared with the case where this structure is not provided, the position of the streak defect can be easily grasped with a simple structure.

  Further, the image evaluation apparatus 10 according to the first embodiment has a configuration in which a warning is given by providing a specific support mark at a position where a streak defect in the printed material 16 to be inspected is extended in a specific direction. Thereby, compared with the case where this structure is not provided, the position of the streak defect can be more easily grasped with a simple structure.

  Further, the image evaluation apparatus 10 according to the first embodiment has a configuration that determines the type of the specific support mark according to the visibility of the streak defect. Thereby, compared with the case where it does not have this structure, the magnitude | size of the visibility of a stripe defect can be grasped | ascertained easily.

  In addition, the image evaluation apparatus 10 according to the first embodiment is predetermined for the display unit 86 when it is determined that the streak defect whose visibility is evaluated by the evaluation unit 22 is a visible streak defect. It is configured to warn by displaying information. Thereby, compared with the case where this configuration is not provided, it is possible to easily grasp that the inspection target printed matter 16 is printed with the inspection target image having an unacceptable streak defect.

  Further, in the image evaluation apparatus 10 according to the first embodiment, as the predetermined information displayed on the display unit 86, the identification for specifying the inspection target printed matter including the streak defect determined to be a visible streak defect Adopt information. Thereby, compared with the case where this configuration is not provided, it is possible to specify the inspection target printed matter 16 on which the inspection target image having unacceptable streak defects is printed with a simple configuration.

  Further, in the image evaluation device 10 according to the first embodiment, as the predetermined information displayed on the display unit 86, the inspection target image having the streak defect and the specification that supports the specification of the position of the streak defect in the inspection target image Support images are used. Thereby, compared with the case where this structure is not provided, the position of the streak defect can be easily grasped with a simple structure.

  In the first embodiment, the aspect in which the evaluation unit 22 performs the color conversion and the visual filter processing is exemplified, but the disclosed technique is not limited to this. Moreover, in the said 1st Embodiment, although the aspect which cuts out a streak area | region by the extraction part 20 was illustrated, the technique of an indication is not limited to this. That is, instead of the extraction unit 20 illustrated in FIG. 3, as an example, the extraction unit 20 </ b> A illustrated in FIG. 22 is applied, and the conversion unit 36 is removed from the evaluation unit 22. The extraction unit 20A illustrated in FIG. 22 is different from the extraction unit 20 illustrated in FIG. 3 in that a subtraction unit 28A is applied instead of the subtraction unit 28, and a conversion unit 26 is interposed between the resolution adjustment unit 26 and the subtraction unit 28A. The difference is that the striped region cutout portion 20 is removed. 22 is different from the extracting unit 20 shown in FIG. 3 in that a specific direction image extracting unit 32A is applied instead of the specific direction image extracting unit 32.

  In FIG. 22, the conversion unit 27 performs color conversion and visual filter processing on the sample image indicated by the sample image information and the inspection target image whose resolution is adjusted. The subtractor 28A calculates a difference value between the sample image information indicating the sample image subjected to the color conversion and the visual filter processing by the conversion unit 27 and the inspection target image information. That is, a difference value is calculated for the density between the sample image that has undergone color conversion and visual filter processing by the conversion unit 27 and the inspection target image. The specific direction image extraction unit 32A extracts a streak defect in a specific direction from the inspection target image that has undergone color conversion and visual filter processing by the conversion unit 27, based on the difference value calculated by the subtraction unit 28.

  In the first embodiment, based on the frequency-resolved image and the direction-resolved image obtained by decomposing the sample image indicated by the evaluation target image and the sample image information for each of a plurality of spatial frequencies and each of a plurality of directions. Although the case where the visibility of the streak defect is evaluated using the mask effect visual model is illustrated, the present invention is not limited to this, for example, based on the frequency-resolved image and the direction-resolved image, the mask effect visual model for a specific direction The visibility of the streak defect may be evaluated. Thereby, compared with the case where it does not have this structure, the evaluation precision of the visibility of a streak defect can be improved with a simple structure.

  Moreover, in the said 1st Embodiment, although the evaluation part 22 evaluated the quality of visibility for the evaluation object image shown by the evaluation object image information produced | generated by the 1st image synthetic | combination part 34 as an object, The disclosed technique is not limited to this. For example, the quality of the visibility may be evaluated by the evaluation unit 22 for a part of the evaluation target image. In this case, for example, an extraction unit 20B illustrated in FIG. 23 may be applied instead of the extraction unit 20 illustrated in FIG. 23 is different from the extraction unit 20 shown in FIG. 3 in that a subtraction unit 28B is applied instead of the subtraction unit 28, and the filter processing unit 250 between the resolution adjustment unit 26 and the subtraction unit 28. The difference is that the candidate area calculation unit 252 and the candidate area cutout unit 254 are interposed. The filter processing 250 performs visual filter processing on the inspection target image whose resolution has been adjusted and the sample image indicated by the sample image information.

  The candidate area calculation unit 252 has a density difference, a color space difference, and a light quantity difference between the inspection target image and the sample image, each of which has been subjected to visual filter processing, for the area specified according to the foresight information from the inspection target image and the sample image. And calculating at least one of a luminance difference, a device signal value difference, and a reflectance. The “foreseeing information” here refers to information for specifying the position of an area predicted in advance as an area where a streak pattern is generated, for example. The candidate area cutout unit 254 compares the calculation result (for example, at least one of density difference, color space difference, light amount difference, luminance difference, light amount difference, and device signal value difference) by the candidate region calculation unit 252 with a threshold value. Then, an area exceeding the threshold is cut out from the inspection target image and the sample image. The subtraction unit 28B calculates a difference value between the inspection target image information and the sample image information for the region cut out by the candidate region cutting unit 254. That is, a difference value for the density between the sample image and the inspection target image is calculated. By having this configuration, the evaluation unit 22 evaluates the visibility of the region cut out by the candidate region cutout unit 254. Thereby, compared with the case where it does not have this structure, the fall of evaluation accuracy can be suppressed further, reducing the load concerning a process.

  Here, at least one of the density difference, color space difference, light amount difference, luminance difference, device signal value difference, and reflectance difference between the inspection target image and the sample image is calculated by the candidate area calculation unit 252. Based on the calculation result, the candidate region cutout unit 254 cuts out the region, but the disclosed technique is not limited to this. For example, the cutout region by the candidate region cutout unit 254 may be determined from the nozzle check pattern. In this case, first, the printing apparatus 12 is caused to print the nozzle check pattern in the blank area of the inspection object printed matter 16. Next, the nozzle check pattern is read by the reading unit 18, and based on the image information obtained by reading the nozzle check pattern by the reading unit 18, the candidate area calculation unit 252 determines the position of the defective nozzle (for example, ink ejection failure) ( For example, the nozzle number) is predicted. Then, the candidate area cutout unit 254 cuts out an area that avoids the area corresponding to the nozzle position predicted by the candidate area calculation unit 252.

  Moreover, in the said 1st Embodiment, although the determination information was illustrated as an example of reference information, the technique of an indication is not limited to this. For example, use-specific determination information that indicates whether or not visual recognition is possible for each predetermined application may be applied as the reference information. The “use” mentioned here refers to, for example, a large format poster, a photo book, a photo book, a catalog, a flyer, and the like. When the usage information indicating the usage of the inspection target printed matter 16 is input, the warning necessity determination unit 44 refers to the usage-specific determination information corresponding to the input usage information, and the visibility indicated by the visibility strength information. It is determined whether or not the intensity is the intensity that requires a warning.

  Moreover, you may apply the determination information classified by observation distance which shows whether it can visually recognize for every observation distance defined beforehand as reference information. Here, the “observation distance” refers to, for example, a distance from the inspection target printed matter 16 to the observer when the inspection target printed matter 16 is actually observed. When the observation distance about the inspection target printed matter 16 is input, the warning necessity determination unit 44 refers to the determination information for each observation distance corresponding to the input observation distance, and the visibility information indicated by the visibility strength information It is determined whether the intensity is an intensity that requires a warning.

  Note that the output resolution of the printing apparatus 12 is preferably an output resolution that is higher than human visual performance as the observation condition. For example, when the observation distance is about 30 cm, it is preferable to employ an output resolution of 350 dpi or more. For example, when the observation distance is 100 cm or more, the output resolution of the printing apparatus 12 is preferably an output resolution of 105 dpi or more.

  The scanner 94 preferably adopts a reading resolution determined according to the output resolution of the printing apparatus 12. For example, when the output resolution of the printing apparatus 12 is 1200 dpi, it is preferable to adopt a reading resolution of 400 dpi or more, which is 1/3 or more of the output resolution of the printing apparatus 12, as the reading resolution of the scanner 94.

  Further, the scanner 94 may adopt a reading resolution determined according to the minimum size of ink droplets used for printing in the printing apparatus 12. For example, when the minimum size of ink droplets used for printing in the printing apparatus 12 is 30 μm, the reading resolution of the scanner 94 is preferably 400 dpi or more.

  Moreover, you may apply the determination information classified by content which shows whether it can visually recognize according to the predetermined content as reference information. Here, “content” refers to, for example, the position of a face image indicating the face of a person in the inspection target image and the skin color of the person. It can be detected by utilizing a face detection function and a skin detection function which are existing technologies. When the content information indicating the content is input, the warning necessity determination unit 44 refers to the content-specific determination information corresponding to the input content information, and the visibility strength indicated by the visibility strength information gives a warning. It is determined whether the strength is required.

  Further, in the first embodiment, an example is given in which all the sample printed matter 14 printed by the printing device 12 is accommodated in the standby area, and all the sample printed matter 14 accommodated in the standby area is read by the scanner 94. However, the disclosed technique is not limited to this. For example, a part of the sample printed matter 14 printed by the printing apparatus 12 may be accommodated in the standby area. For example, when a plurality of recording sheets are held and conveyed on the outer peripheral surface of the drum in the printing apparatus 12, a sample obtained by recording an image on a recording sheet (recording sheet with a specific period) held in a specific holding area. Only the printed material 14 is selected by a sorter (a sorter different from the sorter 96). The selected sample printed matter 14 is accommodated in the standby area by the sorter, and only the sample printed matter 14 accommodated in the standby area is set as the object to be read. In addition to this, a mode in which only the sample printed matter 14 obtained after the trial printing by the printing device 12 a predetermined number of times is accommodated in the standby area can be exemplified.

  In the first embodiment, the specific direction is the conveyance direction of the recording paper, but the disclosed technique is not limited to this. For example, in the case where a printing apparatus that divides the inspection target image into a plurality of lines and records the lines on a recording sheet instead of the printing apparatus 12 is employed, the specific direction may be the main scanning direction in the printing apparatus. . Thereby, compared with the case where this structure is not provided, the streak defect of a specific direction can be extracted with high precision. In this case, it is preferable that the printing method of the printing apparatus is a shuttle scan method, whereby a streak defect in a specific direction can be extracted with higher accuracy.

  Further, in the first embodiment, the case where the exclusion unit 25 is provided in the reading unit 18 is illustrated, but the disclosed technique is not limited to this, and the exclusion unit 25 is removed from the reading unit 18 and excluded. A function equivalent to that of the unit 25 may be assigned to the subsequent extraction unit 20.

  In the first embodiment, the example in which the visibility of the streak defect is evaluated using the mask effect visual model has been described. However, the disclosed technique is not limited to this and is prepared in advance. The visibility of the streak defect may be evaluated using the profile obtained. In this case, for example, the specific component decomposing unit 38 and the second image synthesizing unit 40 are unnecessary, and instead of the visual model evaluation performing unit 42, a profile evaluation performing unit that evaluates the visibility of the streak defect using a profile is used. Applied. Examples of the streak defect evaluation process (profile process) using a profile include a process of integrating a streak defect image in a specific direction. In the profile processing, if there is a portion where the integral value is equal to or greater than a preset threshold value, that portion is detected as a streak. Furthermore, the width exceeding the threshold is also taken into consideration.

  Moreover, in the said 1st Embodiment, although the case where a streak defect was pinpointed directly from a test object image was illustrated, the technique of an indication is not limited to this. For example, a nozzle check pattern may be formed in a blank area (for example, a blank area on the outer peripheral edge) of the inspection target printed matter 16, and a candidate area for specifying a streak defect may be determined using the formed nozzle check pattern. In addition, a chart is created by extracting an area preliminarily assumed as an area where a streak defect is formed from the inspection target image, and recording the extracted area in a blank area at the same position in the main scanning direction. The candidate area may be determined using. Alternatively, a nozzle check pattern may be printed in a blank area of the inspection target print 16, and the candidate area may be narrowed down from the nozzle check pattern and read by the reading unit 18.

  In the first embodiment, the processing for aligning the resolution in the specific direction of the inspection target image indicated by the inspection target image information with the resolution in the main scanning direction is not necessarily performed. In this case, Visual filtering may be performed using a vertical / horizontal deformation filter.

  Moreover, although the said 1st Embodiment gave and demonstrated the example in the case of implement | achieving the extraction part 20, the evaluation part 22, and the warning part 24 with a software structure using the computer 60, the technique of an indication is limited to this. Is not to be done. For example, at least one of the extraction unit 20, the evaluation unit 22, and the warning unit 24 may be realized by a hardware configuration. In this case, there is an example in which an ASIC (Application Specific Integrated Circuit) or a programmable logic device, which is an integrated circuit in which a plurality of function circuits are integrated into one, is used. Further, the extraction unit 20, the evaluation unit 22, and the warning unit 24 may be subjected to distributed processing. The distributed processing may be realized by a combination of a hardware configuration and a software configuration. Further, at least one of the extraction unit 20, the evaluation unit 22, and the warning unit 24 may be realized by a software configuration using a plurality of computers (including peripheral devices). In addition, at least one of the extraction unit 20, the evaluation unit 22, and the warning unit 24 may be realized by cloud computing or grid computing.

  [Second Embodiment]

  In the first embodiment, the aspect in which the visibility of the streak defect is evaluated using the mask effect visual model is exemplified. However, in the second embodiment, the visibility of the streak defect is evaluated by a statistical method. The case will be described. In the second embodiment, the same components as those in the first embodiment are denoted by the same reference numerals, and the description thereof will be omitted. Only parts different from those in the first embodiment will be described.

  FIG. 24 shows an example of the main functions of the evaluation unit 22A according to the second embodiment. As shown in FIG. 24, the evaluation unit 22A differs from the evaluation unit 22 shown in FIG. 4 in that the specific component decomposition unit 38, the second image synthesis unit 40, and the visual model evaluation execution unit 42 are removed. Further, the evaluation unit 22A is different from the evaluation unit 22 shown in FIG. 4 in that an acquisition unit 262, a conversion unit 264, a calculation unit 266, a DB update unit 268, and a statistical evaluation execution unit 272 are provided. An image DB 260 is connected to the acquisition unit 262, and a statistical model DB 270 is connected to the DB update unit 268 and the statistical evaluation execution unit 272.

  In the second embodiment, an example in which the first image composition unit 34 is included in the evaluation unit 22A in FIG. 24 is illustrated, but the first image composition unit 34 is not necessarily included in the evaluation unit 22A. In this case, a streak defect image corresponding to a difference value between the sample image and the inspection target image may be used instead of the evaluation target image obtained by the first image synthesis unit 34.

  In the second embodiment, the evaluation target image and the sample image are adopted as the targets to be decomposed by the specific component decomposition unit 38. However, the present invention is not limited to this, and only the evaluation target image may be set as the target to be decomposed.

  The image DB 260 stores sample image information and inspection target image information that have been used for evaluation by the image evaluation apparatus 10 in the past. FIG. 25 schematically shows an example of the configuration of the image DB 260. As shown in FIG. 25, the image DB 260 stores sample image information and inspection target image information. In addition, the image DB 260 includes subjective streak intensity information indicating the result of subjective evaluation of the strength of the streak defect included in the inspection target image indicated by the inspection target image information (10 levels in which 0 to 9 are indicated). Is stored. Further, the subjective stripe strength information is associated with each inspection target image. Here, in order to avoid complications, one test subject who subjectively evaluates the strength of the streak defect is assumed to be a single subject, and each of a plurality of inspection target images having different streak defects with respect to this test subject is previously stored. The degree of visibility is evaluated for each inspection target image. Here, in order to avoid complications, the inspection target image used as the evaluation target image employs the inspection target image included in the inspection target print 16 that has been an evaluation target in the past.

  The acquisition unit 262 acquires sample image information and inspection target image information from the image DB 260. The conversion unit 264 performs color conversion and visual filter processing on the sample image and the inspection target image indicated by the sample image information and the inspection target image information acquired by the acquisition unit 262, respectively. The calculation unit 266 calculates a difference value between the sample image obtained by performing color conversion and visual filter processing and the inspection target image. The DB update unit 268 updates the stored contents of the statistical model DB 270.

  FIG. 26 schematically illustrates an example of the configuration of the statistical model DB 270. As shown in FIG. 26, the statistical model DB 270 stores the difference value calculated by the calculation unit 266 and the corresponding subjective streak intensity information stored in the image DB 260 in association with each other. A threshold value is associated with each difference value. Here, the “threshold value” is a threshold value that is compared with the density value of each pixel of the evaluation target image indicated by the evaluation target image information. If this threshold value is exceeded, it is evaluated that the streak pattern is visible. Is done. Note that the “threshold” mentioned here is not an essential component for realizing the disclosed technology. Accordingly, the disclosed technique is established even if there is no “threshold”.

  The statistical evaluation execution unit 272 refers to the statistical model DB 270 and uses a statistical method to determine the visibility of the streak defect of the evaluation target image obtained by performing the color conversion and the visual filter processing by the conversion unit 36. evaluate. In the second embodiment, for example, the statistical evaluation execution unit 272 first performs a sample image and a streak based on an evaluation target image and a sample image obtained by performing color conversion and visual filter processing by the conversion unit 36. A difference value from the defect image is calculated for each pixel, and an average value of the calculated difference values is calculated. Next, out of the threshold values stored in the statistical model DB 270, the threshold value associated with the difference value corresponding to the calculated average value is read, and the difference value calculated for each pixel is compared with the threshold value. Then, the pixel corresponding to the difference value exceeding the threshold value is evaluated as a pixel constituting a visible streak pattern, and evaluation result information including a comprehensive evaluation result and subjective streak intensity information for all pixels is generated. To do. In the second embodiment, a value corresponding to the average value of the difference values calculated for each pixel is adopted as the threshold value, but the disclosed technique is not limited to this. For example, a value corresponding to a representative value of the calculated difference value for each pixel (for example, a statistical value other than the average value such as the mode value or the median value of the calculated difference value for each pixel) may be adopted as the threshold value. Good.

  In the second embodiment, an example in which the statistical evaluation execution unit 272 calculates the difference value between the sample image and the streak defect image for each pixel based on the evaluation target image and the sample image will be described. However, without generating the evaluation target image, the streak defect image information generated by the specific direction image extraction unit 32 is acquired, and the difference between the streak defect image and the sample image indicated by the streak defect image information is obtained. A value may be calculated for each pixel.

  FIG. 27 shows an example of the configuration of the electrical system of the image evaluation apparatus 10A according to the second embodiment. The image evaluation apparatus 10A shown in FIG. 27 differs from the image evaluation apparatus 10 shown in FIG. 6 in that an image evaluation processing program 70A is applied instead of the image evaluation processing program 70. Compared to the image evaluation apparatus 10 shown in FIG. 6, the image evaluation apparatus 10 </ b> A has a point that the DB update processing program 274 is stored in the storage unit 66, a point that the image DB 260 is stored in the storage unit 66, and statistics Another difference is that the model DB 270 is stored.

  The image evaluation processing program 70A differs from the image evaluation processing program 70 shown in FIG. 6 in that an evaluation process 78A is applied instead of the evaluation process 78, and the CPU 62 executes the evaluation process 78A. , Operates as an evaluation unit 22A shown in FIG.

  Next, as an operation of the second embodiment, a DB update process performed by the image evaluation apparatus 10A when the CPU 62 executes the DB update process program 274 will be described with reference to FIG. In the DB update processing shown in FIG. 28, first, in step 300, the acquisition unit 262 obtains sample image information, inspection target image information corresponding to the sample image information, and subjective streak intensity information corresponding to the inspection target image information. It is determined whether or not the image has been input to the image DB 260. If the sample image information, the inspection target image information, and the subjective streak intensity information are not input in step 300, the determination is denied and the determination in step 300 is performed again. If sample image information, inspection object image information, and subjective streak intensity information are input in this step 300, the determination is affirmed and the routine proceeds to step 302.

  In step 302, the acquisition unit 26 acquires the latest stored sample image information and inspection target image information from the image DB 260, and then proceeds to step 304. In step 304, the conversion unit 264 performs color conversion and visual filter processing on the sample image and the inspection target image indicated by the sample image information and the inspection target image information acquired in step 302, respectively.

  In the next step 306, the calculation unit 266 calculates a difference value between the sample image that has been subjected to the color conversion and the visual filter conversion in step 304 and the inspection target image, and then the process proceeds to step 308. In step 308, the DB update unit 268 acquires the latest subjective streak subjective information from the image DB 260, and then the process proceeds to step 310. In step 310, the difference value calculated in step 306, the threshold value uniquely determined according to the difference value, and the subjective streak intensity information acquired in step 308 are stored in the statistical model DB 270 by the DB update unit 268. Remembered. Thereby, the statistical model DB 270 is updated. And after the process of this step 310 is performed, this DB update process is complete | finished.

  Next, image evaluation processing performed by the image evaluation apparatus 10A when the CPU 62 executes the image evaluation processing program 70A will be described. Here, a process different from the image evaluation process described in the first embodiment will be described, and a description of the same process as that in the first embodiment will be omitted. As shown in FIG. 14, the image evaluation process according to the second embodiment is different from the image evaluation process described in the first embodiment in that step 204A is applied instead of step 204. In step 204A, evaluation processing is performed by the evaluation unit 22A, and then the process proceeds to step 206.

  FIG. 29 shows an example of the flow of evaluation processing according to the second embodiment. The evaluation process according to the second embodiment shown in FIG. 29 is different from the evaluation process shown in FIG. 19 in that steps 320 to 332 are applied instead of steps 204C to 204E. Hereinafter, steps different from the steps included in the evaluation process shown in FIG. 19 will be described, and the same steps will be denoted by the same step numbers and the description thereof will be omitted. Here, a case will be described in which a streak defect image is included in the evaluation target image in order to avoid complications.

  In the evaluation process shown in FIG. 29, in step 320, the sample image and the streak of the target pixel of the streak defect image included in the evaluation target image obtained by performing the conversion process in step 204B by the statistical evaluation execution unit 272. A difference value with respect to the density of the defect image is calculated. In the next step 322, the statistical evaluation execution unit 272 stores the difference value calculated in step 320 in the storage unit 66 for each pixel.

  In the next step 324, the statistical evaluation execution unit 272 determines whether or not the processing in steps 320 and 322 has been completed for all pixels in the streak defect image. If the processing in steps 320 and 322 is not completed for all pixels in the streak defect image in step 324, the determination is negative and the processing proceeds to step 320. In step 324, when the processing in steps 320 and 322 is completed for all pixels in the streak defect image, the determination is affirmed and the process proceeds to step 326.

  In step 326, the statistical evaluation execution unit 272 calculates an average value for the difference value for each pixel stored in the storage unit 66 in step 320. In the next step 328, the statistical evaluation execution unit 272 acquires a threshold value associated with the difference value corresponding to the average value calculated in step 326 from the statistical model DB 270. In addition, as a difference value corresponding to the average value calculated in the above step 326, if a difference value that completely matches the calculated average value is not stored in the statistical model DB 270, the difference value calculated in the above step 320 A difference value that most closely approximates the average value of these may be used instead. When the average value of the calculated difference values is the median value of the two difference values stored in the statistical model DB 270, for example, the smaller one of the two difference values is calculated in the above step 320. What is necessary is just to employ | adopt as a difference value equivalent to the average value of the obtained difference value. Conversely, the larger difference value of the two difference values may be adopted as the difference value corresponding to the average value of the difference values calculated in step 320.

  In the next step 330, the statistical evaluation execution unit 272 acquires the difference value for the target pixel from the difference value for each pixel stored in the storage unit 66 in step 322. In the next step 332, the statistical evaluation execution unit 272 determines whether or not the difference value acquired in step 330 is greater than the threshold acquired in step 328. In step 332, if the difference value acquired in step 330 does not exceed the threshold acquired in step 328, the determination is negative and the process proceeds to step 338. In step 332, if the difference value acquired in step 330 is greater than the threshold acquired in step 328, the determination is affirmed and the process proceeds to step 334.

  In step 334, the statistical evaluation execution unit 272 acquires subjective streak intensity information corresponding to the threshold acquired in step 328 from the statistical model DB 270. The acquired subjective streak intensity information is stored in the storage unit 66 in association with the difference value acquired in step 330.

  In the next step 336, the statistical evaluation execution unit 272 determines whether or not the processing in steps 330 to 334 has been completed for all the difference values for each pixel stored in the storage unit 66 in step 322. . In step 336, if the processing of steps 330 to 334 has not been completed for all the difference values for each pixel stored in the storage unit 66 in step 322, the determination is negative and the process proceeds to step 330. In step 336, when the processing of steps 330 to 334 has been completed for all the difference values for each pixel stored in the storage unit 66 in step 322, the determination is affirmed and the process proceeds to step 338.

  In step 338, the statistical evaluation execution unit 272 evaluates the visibility of the streak defect included in the evaluation target image, and generates evaluation result information indicating the evaluation result. In this step 338, for example, when subjective streak intensity information is stored in the storage unit 66, all the subjective streak intensity information stored in the storage unit 66 is comprehensively evaluated and included in the evaluation target image. The visibility of streak defects is evaluated. For example, a subjective value indicating a numerical value (for example, 8) that exceeds a predetermined value (for example, 7) among the magnitude of the numerical value for the 10-level evaluation, the number of subjective line strength information, or the numerical value for the 10-level evaluation, which is indicated by the subjective line strength information Evaluation is performed based on the distribution of pixels associated with the stripe intensity information. The evaluation result information includes information used for the evaluation (for example, subjective streak intensity information and information indicating the pixel distribution) in addition to the evaluation result of the visibility of the streak defect. Is used in the evaluation by the warning necessity determination unit 44 of the warning unit 24 in the subsequent stage.

  As described above, the image evaluation apparatus 10A according to the second embodiment has a configuration for evaluating the visibility of the extracted streak image using a statistical method. Compared with the case where this configuration is not provided, it is possible to improve the evaluation accuracy of the visibility of the streak defect.

  In the second embodiment, the example of evaluating the visibility of the streak defect with reference to the statistical model DB 270 has been described. However, the disclosed technique is not limited to this, and a multiple regression equation is used. It may be used to evaluate the visibility of streak defects. In this case, instead of the evaluation unit 22A, an evaluation unit 22B shown in FIG. 30 may be used as an example. 30 differs from the evaluation unit 22A illustrated in FIG. 24 in that the statistical evaluation execution unit 272A is applied instead of the statistical evaluation execution unit 272, and the DB update unit 268 and the statistical model DB 270 are replaced. And the regression formula update unit 340 is applied.

  The regression equation update unit 340 holds the multiple regression equation provided to the statistical evaluation execution unit 272A and updates the regression coefficient of the multiple regression equation based on the latest data. The “multiple regression equation” here uses, for example, a numerical value of 10-level evaluation indicated by the difference value calculated by the calculation unit 266 and the subjective streak intensity information stored in the image DB 260 the latest. In addition, a numerical value (for example, a numerical value corresponding to a numerical value of 10-level evaluation indicated by the subjective streak intensity information) indicating the visibility of the streak defect included in the evaluation target image is used as an objective variable.

  The statistical evaluation execution unit 272A evaluates the visibility of the streak defect included in the evaluation target image using the multiple regression equation currently held by the regression equation update unit 340, and shows the evaluation result Generate result information. The “evaluation result information” mentioned here includes, for example, a numerical value (solution of multiple regression equations) indicating the quality of the visibility of the streak defect included in the evaluation target image.

  Thus, by evaluating the visibility of the streak defect using the multiple regression equation, the evaluation accuracy of the visibility of the streak defect can be improved with a simple configuration.

  The disclosed technique is not limited to the multiple regression equation, and for example, SIMM, PCA, SVR, AdaBoost, etc. may be adopted.

  In the second embodiment, the case where the subject who subjectively evaluates the strength of the striated defect is exemplified, but the disclosed technique is not limited to this, and the subject may be a plurality of subjects. . In this case, an average value, a mode value, and the like of the numerical value representing the subjective streak intensity, which is an evaluation result by a plurality of subjects, may be stored in the image DB 260.

  In the second embodiment, the evaluation target image to be visually recognized by the subject is the inspection target image included in the inspection target printed matter 16 subjected to the past evaluation. However, the present invention is not limited to this, and an image similar to the inspection target image ( It may be an image that has never been actually used as an inspection target image.

  [Third Embodiment]

  In the first embodiment, the warning unit 24 refers to the reference information in the reference information DB 46 to determine whether or not the visibility strength of the streak defect is a strength that requires a warning. The third embodiment Then, it is determined whether or not the visibility strength is a strength that requires a warning according to the characteristics of the customer. In the following description, the same constituent members as those described in the first embodiment are denoted by the same reference numerals, description thereof is omitted, and portions different from those in the first embodiment are described.

  FIG. 31 shows an example of main functions of the warning unit 24A according to the third embodiment. The warning unit 24A shown in FIG. 31 differs from the warning unit 24 shown in FIG. 5 in that a warning necessity determination unit 44A is applied instead of the warning necessity determination unit 44. In the first embodiment, the reference information DB 46 is connected to the warning necessity determination unit 44. However, in the third embodiment, a customer characteristic DB 46A is applied instead of the reference information DB 46.

  To the warning necessity determination unit 44A, customer identification information for identifying a customer and usage information indicating the usage of the printed material 16 to be inspected (for example, large format poster, photo book, photo book, catalog, flyer, etc.) are input. The warning necessity determination unit 44A refers to the customer characteristic DB 46A from the input customer identification information and usage information, and determines whether or not the visibility strength of the streak defect is a strength that requires a warning.

  The customer characteristic DB 46A is a data group in which information (hereinafter referred to as “customer characteristic information”) such as order contents and past transaction results (history) is stored for each customer who is an orderer (order owner). In the customer characteristic DB 46A of this example, for each customer, for each use of printed matter and content type, the customer describes the image quality approved by the customer in the past (acknowledged as OK) with the numerical value of the index “image quality index”. Characteristic information is stored. The image quality index is an index that defines the image quality of a printed material, and is defined as a numerical value including an evaluation of at least one type, preferably a plurality of types of image quality attributes, associated with a physical quantity that can be actually measured from the printed material. A specific example of the image quality index will be described later. In this specification, the image quality index is represented by the symbol “Q”.

  FIG. 32 is an explanatory diagram of the image quality index. The image quality index in this example is an index that defines the overall image quality of a printed material, and is defined as a value that combines evaluations of a plurality of image quality attributes. That is, the overall image quality can be represented by a tree structure of a plurality of image quality attributes, and is grasped as a quality that comprehensively considers each attribute. In the example of FIG. 32, six types of image quality attributes are shown as image quality attributes: image noise, color gradation, sharpness, character property, gloss, and non-uniformity. Some of the six types of image quality attributes exemplified here may be used, and other image quality attributes may be added.

  Each attribute is associated with one or more of the print quality items, and the image quality attribute and the print quality item correspond to physical quantities that can actually be measured from the print results (corresponding to “physical measurement values” in FIG. 32). , Described as “image quality measurement value”).

  FIG. 33 is a conceptual diagram showing how to obtain the image quality index. As a reference for explaining the concept shown in FIG. 33, there is “Peter G. Engeldrum,“ Psychometric Scaling: a toolkit for imaging systems development ”, pp. 8-17., 2000.

  The values of physical technology variables are obtained by printing a test chart or an actual image and measuring the printing result. The measured technical variable is represented by a symbol “X”, and specific values are represented by X1, X2, X3.

  A value X obtained by measurement such as chart measurement and a physical image parameter (physical quantitative value) Yi as an image quality measurement value are associated according to a predetermined system model. As the system model, for example, ISO / IEC 13660 relating to a method for measuring hard copy image quality can be applied. A physical image parameter (physical quantitative value) Yi = Fi (X) as an image quality measurement value is obtained from a value X obtained by chart measurement or the like according to a function Fi (X) defined by the system model.

  The physical image parameter referred to here is a physical evaluation value (index) generally called “˜degree”, such as RMS (root mean square granularity) granularity. There can be a plurality of types of such physical image parameters.

  The customer perception (customer perception) corresponding to the image quality attribute or the print quality item is a sensory evaluation value generally called “˜sex” such as graininess. This sensory evaluation value (represented by Zi in the figure) is represented as a numerical value that does not depend on the application.

  The sensory evaluation value Zi representing the image quality attribute (or print quality item) is calculated according to a function Gi (Y) defined by the visual algorithm (Zi = Gi (Y)).

  The customer quality preference (customer image quality preference) corresponding to the overall print image quality is a sensory evaluation value of the overall image quality combining a plurality of image quality attributes. This total image quality is defined for each application. The overall image quality evaluation value Qi is calculated by a function Hi (Z) defined in the image quality model, and a different function Hi (Z) is defined for each application.

  As a simple example, a linear model may be used as each model or algorithm for associating the physical quantity X, the image quality measurement value Y, the image quality attribute (or print quality item) Z, and the print overall image quality Q obtained by actual measurement. More specifically, a specific function is determined through an evaluation experiment or the like.

  Correspondence relationships (functions indicated by Fi, Gi, Hi) between the sensory evaluation values Qi, Zi and the physical quantitative values (physical evaluation values) X, Y are obtained in advance. At that time, as for the relationship between the print overall image quality Qi and the image quality attribute / print quality item (Zi), the model parameter is optimized for each use of the printed matter. For example, in applications such as a “photo book” with many photographs, image quality attributes relating to images are emphasized. On the other hand, in applications such as “document” with many characters, image quality attributes relating to characters are emphasized.

  The calculation of the print overall image quality Qi may be a linear model of multiple regression analysis, a linear model with a weighting factor corresponding to the area ratio of the content type, or a non-linear model. Thus, according to the function Hi defined for each application, the overall image quality Q1 for catalogs, the overall image quality Q2 for photo books, the overall image quality Q3 for flyers, and the like are calculated.

  In the third embodiment, an overall image quality obtained by comprehensively evaluating a plurality of image quality attributes is used as the image quality index. However, instead of the overall image quality, individual image quality attributes before comprehensive evaluation, or the appropriate ones of these. A combination of these may be used.

  FIG. 34 schematically shows an example of the configuration of the customer characteristic DB 46A. As shown in FIG. 34, information (customer name, customer ID, etc.) for identifying a customer is stored in the customer characteristic DB 46A. The customer characteristic DB 46A stores, as customer characteristic information, an image quality index that satisfies the customer's required quality for each application and content type conditions, and an image quality index that has not yet reached the customer's required quality. In the following, for convenience of explanation, an image quality index that satisfies the customer's required quality may be expressed as “OK_Q”, and an image quality index that has not reached the customer's required quality may be expressed as “NG_Q”.

  As a specific example of “OK_Q”, there is an image quality index in which customer's approval / approval (OK judgment) is obtained for the provided printed matter. As a specific example of “NG_Q”, there is an image quality index in which the provided printed matter cannot satisfy the required quality of the customer and rejection / non-approval (NG judgment) is expressed by the customer.

  Such data can be accumulated as history information from the results of deliveries for past orders. That is, through the transaction with the customer, the image quality index of the provided printed matter and the customer's evaluation (OK or NG) for the printed matter are associated with each other and stored in the customer characteristic DB as history data.

  Since the quality required for the printed material varies greatly depending on its use, the customer characteristic DB 46A classifies data from the viewpoint of “use”. The printed image (one picture) may include various contents such as a photo part, a character part, a line drawing part, and an illustration part. The viewpoint of classifying what the elements (contents) constituting the image content of the printed material are is classification by “content type”.

  There is a relationship in which there is a more detailed subcategory called “content type” in the major category “use”. In the third embodiment, image quality index data is stored for a combination of usage and content type.

  Other items are not essential, but information on items that are considered to have high utility value is registered in the customer characteristic DB. In the example shown in FIG. 34, a job ID (identification code) for identifying a print job (Job) for each order, an order reception date, a cost (denoted by a symbol “C”), and a delivery date (sign “D” ”.) That is, as a job history in past transactions, whether the customer is satisfied (OK) or satisfied with respect to the image quality (Q), cost (C), and delivery date (D) of the printed matter provided according to the order Information indicating whether or not there was (NG) is accumulated in the customer characteristic DB 46A.

  In this example, the overall image quality is used as the image quality index Q, and the evaluation values (Zi; i = a, b, c) of image quality attributes (here, six types of attributes a to f) included as evaluation items for the overall image quality. , d, e, f) are also recorded. Specifically, for example, six types of attributes such as noise, color gradation, sharpness, character, gloss, and non-uniformity shown in FIG. 32 correspond to a to f. Although not shown in FIG. 34, a corresponding value is described in each cell of the table.

  In addition, various statistical processes can be performed from such a customer characteristic DB, and by processing various data registered in the customer characteristic DB, preferences and trends for each customer, an average of all customers Secondary information such as an OK level (an average value of OK_Q) can be generated. Such secondary information may be generated from the customer characteristic DB 46A as needed based on an operator's instruction as needed, or may be automatically generated at an appropriate timing periodically or irregularly. . Further, the generated secondary information (DB processing data) may be stored in the customer characteristic DB 46A, registered in another database, or stored in another storage device. is there.

  In the third embodiment, standard parameters such as an average value of a plurality of customers are held in the customer characteristic DB 46A. These standard data are used, for example, to estimate the required quality of new customers.

  Further, in this example, as a clue to guess the customer's required quality, the distribution data (referred to as “OK / NG image quality index distribution”) of the image quality index Q, the number of times of OK, and the number of times of NG as shown in FIG. ) Is generated. Histograms of customer's OK and NG judgments for image quality index Q (total image quality) by use from the accumulated data group of image quality index Q of past delivery and customer evaluation (OK / NG judgment) for the printed matter Can be obtained.

  In FIG. 35, the distribution indicated by a solid line is OK number distribution data (referred to as “OK image quality distribution data”). In FIG. 35, the distribution indicated by the broken line is NG count distribution data (referred to as “NG image quality distribution data”). Such an image quality index distribution can be generated from a group of data of the same use using data of a specific customer, or a plurality of customers (for example, all registered customers or a part of customers). It can also be generated from a data group for the same purpose.

  For example, when dealing with the same customer many times, the distribution of the image quality level (OK_Q) in which the quality of the printed matter is determined as “OK” and the image quality level (NG_Q) in which it is determined as “NG” becomes clear. . For regular customers, an OK / NG image quality index distribution for each customer's usage is generated from the data of the specific customer. For new customers, a distribution is created using data of other customers related to the same application.

  In addition, in the case of a new use not registered in the customer characteristic DB 46A, the distribution is estimated using other use data, or the image quality attribute items emphasized in the new use and the balance between the items are considered. Distribution can be estimated using the assumptions made.

  Furthermore, using the deviation (for example, standard deviation) of the specific customer data from the average data including other customers, the items to be emphasized in the judgment of the preference of the specific customer (regular customer), for example, OK / NG Item) and items that are not so important (light-look items) can be estimated. Based on this estimation, it is possible to propose a printed matter that better suits the customer's print request (order).

  FIG. 36 shows an example of the configuration of the electrical system of the image evaluation apparatus 10B according to the third embodiment. The image evaluation apparatus 10B shown in FIG. 36 differs from the image evaluation apparatus 10 shown in FIG. 6 in that an image evaluation processing program 70B is applied instead of the image evaluation processing program 70. The image evaluation apparatus 10B is different from the image evaluation apparatus 10 shown in FIG. 6 in that a customer DB 46A is stored in the storage unit 66 instead of the reference information DB 46.

  The image evaluation processing program 70B is different from the image evaluation processing program 70 shown in FIG. 6 in that a warning process 80A is applied instead of the warning process 80, and the CPU 62 executes the warning process 80A. , Operates as an evaluation unit 24A shown in FIG.

  Next, image evaluation processing performed by the image evaluation apparatus 10B when the CPU 62 executes the image evaluation processing program 70B will be described. Here, a process different from the image evaluation process described in the first embodiment will be described, and a description of the same process as that in the first embodiment will be omitted. As shown in FIG. 14, the image evaluation processing according to the third embodiment is different from the image evaluation processing described in the first embodiment in that step 206A is applied instead of step 206. In step 206A, an evaluation process is performed by the warning unit 24A, and then the image evaluation process is terminated.

  FIG. 37 shows an example of the flow of warning processing according to the third embodiment. The warning process according to the third embodiment shown in FIG. 37 is different from the warning process shown in FIG. 20 in that steps 350 to 354 are applied instead of steps 206B and 206C. In the following, steps different from the steps included in the warning process shown in FIG. 20 will be described, and the same steps will be denoted by the same step numbers and description thereof will be omitted. Here, a case will be described in which a streak defect image is included in the evaluation target image in order to avoid complications.

  In the warning process shown in FIG. 37, in step 350, the customer identification information and usage information are acquired by the warning necessity determination unit 44A, and then the process proceeds to step 352. In step 352, customer characteristic information is acquired by the warning unit 24A.

  In the next step 354, the customer necessity information acquired in step 352 is referred to by the warning necessity determination unit 44A based on the customer identification information and usage information acquired in step 350. Then, based on the reference result, it is determined whether or not the visibility strength of the streak defect is a strength that requires a warning. In this step 354, for example, the warning necessity determination unit 44A refers to the customer characteristic information corresponding to the customer identification information and the usage information, and the visibility strength indicated by the visibility strength information included in the evaluation result information is determined. It is determined whether or not the intensity requires a warning. If the intensity does not require a warning in step 352, the determination is denied and the warning process is terminated. If it is the intensity that requires a warning in step 352, the determination is affirmed and the routine proceeds to step 206D.

DESCRIPTION OF SYMBOLS 10 Image evaluation apparatus 12 Printing apparatus 14 Sample printed matter 16 Inspection object printed matter 18 Reading part 20,20A, 20B Extraction part 22,22A, 22B Evaluation part 24,24A Warning part 25 Exclusion part

Claims (27)

A storage unit for storing sample image information indicating a virtual sample image assuming a sample image included in a sample printed material recognized as a non-defective product;
A reading unit that reads an inspection target image included in an inspection target printed matter obtained by printing an image indicated by the image information on a recording medium by a printing device based on image information indicating an image corresponding to the sample image; A reading unit that reads the inspection target image such that the resolution in a specific direction is lower than the resolution in a direction intersecting the specific direction;
Based on the difference value between the sample image information stored in the storage unit and the inspection target image information indicating the inspection target image obtained by reading the inspection target image by the reading unit and corresponding to the sample image An extraction unit that extracts a streak defect having a streak pattern in a specific direction from the inspection target image indicated by the inspection target image information, the inspection unit obtained by reading the inspection target image by the reading unit After the resolution in the specific direction of the inspection target image indicated by the target image information is aligned with the resolution in the direction intersecting the specific direction, the inspection is performed based on a difference value between the sample image information and the inspection target image information. An extraction unit for extracting the streak defect from the inspection target image indicated by the target image information;
An evaluation unit that evaluates the visibility of the streak defect extracted by the extraction unit using a mask effect visual model;
An image evaluation apparatus including:   The image evaluation apparatus according to claim 1, wherein the evaluation unit evaluates the visibility of the line defect using the mask effect visual model for the specific direction.   The image evaluation apparatus according to claim 1, wherein the streak defect to be evaluated by the evaluation unit is a streak defect having a spatial frequency corresponding to human visual characteristics.   The image evaluation apparatus according to claim 1, wherein the streak defect to be evaluated by the evaluation unit is a streak defect in a color space corresponding to human visual characteristics.   5. The extraction unit according to claim 1, wherein the extraction unit extracts the streak defect by targeting a predetermined region in an inspection target image indicated by the inspection target image information. Image evaluation device.   The image evaluation apparatus according to claim 5, wherein the predetermined region is a region assumed in advance as a region where the streak defect is formed.   The predetermined area further corresponds to a difference value that exceeds a threshold value with a difference value between the inspection target image indicated by the inspection target image information and the sample image indicated by the sample image information. The image evaluation apparatus according to claim 6, wherein the image evaluation area is an area to be processed.   The image evaluation apparatus according to claim 6, wherein the area assumed in advance is an area specified based on at least one characteristic of the printing apparatus and the reading unit.   The image evaluation apparatus according to claim 8, wherein the characteristics of the printing apparatus include a characteristic of deterioration with time of components in the printing apparatus.   The image evaluation apparatus according to claim 8, wherein the characteristic of the reading unit includes a characteristic of deterioration with time of a constituent member in the reading unit.   The predetermined region is a region corresponding to a difference value that exceeds a threshold value with a difference value between the inspection target image indicated by the inspection target image information and the sample image indicated by the sample image information. The image evaluation apparatus according to claim 5.   The image evaluation apparatus according to claim 1, wherein the specific direction is a conveyance direction of the recording medium in the printing apparatus.   The image evaluation apparatus according to claim 12, wherein a printing method of the printing apparatus is a single pass method.   The said specific direction is made into the main scanning direction in the said printing apparatus, when the said printing apparatus divides | segments the said test object image into a some line, and records it with respect to the said recording medium for every line. The image evaluation apparatus according to any one of 11.   The image evaluation apparatus according to claim 14, wherein a printing system of the printing apparatus is a shuttle scan system. An exclusion unit that excludes the influence of the characteristics of the reading unit from at least the inspection target image information of the sample image information and the inspection target image information;
The extraction unit includes the sample image information obtained by excluding the influence of the characteristics of the reading unit from at least the inspection target image information of the sample image information and the inspection target image information by the exclusion unit and the inspection target. The image evaluation device according to any one of claims 1 to 15, wherein the streak defect is extracted from an inspection target image indicated by the inspection target image information based on a difference value with image information.   17. The warning unit according to claim 1, further comprising a warning unit that warns when it is determined that the streak defect whose visibility has been evaluated by the evaluation unit is a visible streak defect. Image evaluation device.   The image evaluation apparatus according to claim 17, wherein the warning unit issues a warning by directly processing the printed material to be inspected that includes the streak defect that is determined to be a visible streak defect.   The image evaluation apparatus according to claim 18, wherein the warning unit issues a warning by adding a specific support mark that supports specification of a position of the streak defect in the inspection target printed matter to the inspection target printed matter.   The image evaluation apparatus according to claim 19, wherein the warning unit gives a warning by giving the specific support mark to a position where the streak defect in the inspection target printed matter is extended in the specific direction.   The image evaluation apparatus according to claim 19 or 20, wherein the type of the specific support mark is determined according to the visibility of the streak defect.   The warning unit further warns the display unit by displaying predetermined information when it is determined that the streak defect whose visibility has been evaluated by the evaluation unit is a visible streak defect. The image evaluation apparatus according to any one of claims 17 to 21.   The warning section warns by displaying predetermined information on a display section when it is determined that the streak defect whose visibility has been evaluated by the evaluation section is a visible streak defect. The image evaluation apparatus according to 17.   The image evaluation apparatus according to claim 22 or 23, wherein the predetermined information includes specific information that specifies the printed matter to be inspected that includes the streak defect determined to be a visible streak defect. The display unit displays an image,
25. The method according to any one of claims 22 to 24, wherein the predetermined information includes the inspection target image having the streak defect and a specification support image that supports specification of a position of the streak defect in the inspection target image. The image evaluation apparatus described. Based on sample image information stored in a storage unit that stores sample image information indicating a virtual sample image assuming a sample image included in a sample printed material recognized as a non-defective product, and image information indicating an image corresponding to the sample image A reading unit that reads an inspection target image included in an inspection target printed matter obtained by printing an image indicated by the image information on a recording medium by a printing device, and having a resolution in a specific direction intersecting the specific direction A difference value between the inspection target image obtained by reading the inspection target image by a reading unit that reads the inspection target image so as to be lower than the resolution, and indicating an image corresponding to the sample image. An extraction step of extracting a streak defect having a streak pattern in a specific direction from the inspection target image indicated by the inspection target image information, the reading unit The sample image information and the inspection object are obtained after aligning the resolution in the specific direction of the inspection object image indicated by the inspection object image information obtained by reading the inspection object image with the resolution in the direction intersecting the specific direction. Based on a difference value with image information, an extraction step of extracting the streak defect from the inspection target image indicated by the inspection target image information;
An evaluation step for evaluating the visibility of the streak defect extracted by the extraction step using a mask effect visual model;
An image evaluation method including: Computer
The program for functioning as the said extraction part and the said evaluation part in the image evaluation apparatus of any one of Claims 1-25.