JP2008089574A - Image processor, data processor, parameter adjustment method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a data processor, an image processor, a parameter adjustment method, and a program which can adjust parameters concerning data processing accurately with an easy manipulation. <P>SOLUTION: For making the adjustment of parameter values concerning image processing, a target data creating part 140 receives a correction directive to result data D2 from an operator and creates target data D3, then, a parameter adjustment making part 170 detects a parameter value which can make result data D2 agreed with the target data D3 within a predetermined tolerance and determines the value for an adjustment value. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a technique for adjusting parameters (for example, parameters for defect detection processing) related to various data processing (for example, image processing).

  In the printing workflow, an inspection (hereinafter referred to as “plate inspection / printed material inspection”) is performed to check whether an unexpected defect or the like has occurred in the prepared plate-making film, printing plate, or printed material. In this plate inspection / printed material inspection, an image to be inspected (specifically, image data obtained by scanning a printed material to be inspected) is compared with the reference image data from which the image is created. The difference is detected as a defect.

  In plate inspection / printed material inspection, defects occurring in an image to be inspected are detected through various processes (for example, blurring process, swaying process, comparison process, isolated point removal process, etc.). The defect detection sensitivity varies depending on the environmental parameter value of each process. That is, in order to detect a defect with an appropriate sensitivity, it is necessary to adjust the parameter value appropriately.

  In the past, this parameter adjustment work has been performed by a trial-and-error method in which each user corrects it based on empirical rules.

  As a prior art for plate inspection processing, for example, print data to be subjected to plate inspection is divided into regions according to image characteristics, and plate inspection processing is performed using parameters corresponding to the attributes of the image region for each region. The technique to perform is devised (refer patent document 1). Even in this case, it is necessary to appropriately adjust the parameter value in order to perform proper defect detection in each image region.

JP 2004-258470 A

  However, in the plate inspection / printed material inspection, various processes are combined as described above, and there are many types of parameters to be adjusted. Therefore, much time and labor had to be spent adjusting the parameters.

  Even if the parameters are somehow adjusted, if there is a need to change the defect detection sensitivity suddenly, there is no time to adjust the parameter values again. I had to do an inspection by confirmation.

  The present invention has been made in view of the above problems, and is a data processing apparatus and image processing apparatus capable of accurately adjusting one or more parameters relating to data processing (for example, image processing) with an easy operation. It is an object to provide a parameter adjustment method and a program.

  According to the first aspect of the present invention, an image processing means for performing predetermined image processing and a correction instruction for a result image created as a result of the image processing are received, and a target image in which the correction instruction is reflected in the result image is obtained. Target image creating means for creating, and parameter adjusting means for adjusting parameter values related to the image processing so that a result image created as a result of the image processing matches the target image.

  A second aspect of the present invention is the image processing apparatus according to the first aspect, wherein the image processing unit compares the reference image with an inspection target image obtained by outputting the reference image. Image processing is performed to generate a result image in which the detected defective portion is detected and the detected defective portion is displayed in a display form different from that of the non-defective portion. Specify the detection sensitivity of.

  The invention according to claim 3 is the image processing apparatus according to claim 1 or 2, wherein the target image creating means displays the result image on a display screen, and the result is displayed on the display screen. Accepts input of correction instructions for images.

  A fourth aspect of the present invention is the image processing apparatus according to any one of the first to third aspects, wherein the parameter adjusting means obtains a result obtained under each parameter value while changing the value of the parameter. It is determined whether or not the image matches the target image, and the adjustment value of the parameter is determined to be a parameter value obtained as a result image that matches the target image.

  A fifth aspect of the present invention is the image processing apparatus according to the fourth aspect, wherein the parameter adjusting means is configured to include an adjustment value of the parameter from a parameter value obtained as a result image that matches the target image. To the median of the set.

  A sixth aspect of the present invention is the image processing apparatus according to the fourth aspect, wherein the parameter adjusting means changes the parameter adjustment value, and the result image that matches the target image changes the parameter value. To determine the parameter value when first obtained.

  A seventh aspect of the present invention is the image processing apparatus according to any one of the first to third aspects, wherein the parameter adjusting means matches the target image from among one or more parameter value combinations. A combination that gives an image is extracted by a genetic algorithm, and an adjustment value of the parameter is determined as each value constituting the extracted combination.

  The invention according to claim 8 is the image processing device according to any one of claims 4 to 7, further comprising: an adjustment range specifying unit that specifies a range in which the parameter value is changed based on an input set value. Further prepare.

  A ninth aspect of the present invention is the image processing apparatus according to any one of the fourth to eighth aspects, wherein a change amount specification that specifies a change amount when changing the parameter value based on an input setting value. Means.

  A tenth aspect of the present invention is the image processing apparatus according to any one of the first to ninth aspects, wherein an adjustment is made among parameters relating to the image processing based on a predetermined feature amount acquired from the correction instruction. Adjustment target parameter specifying means for specifying a target parameter is provided.

  According to an eleventh aspect of the present invention, there is provided a target image creating step of receiving a correction instruction for a result image created as a result of image processing from an operator and creating a target image in which the correction instruction is reflected in the result image; And a parameter adjustment step of adjusting a parameter value related to the image processing so that a result image created as a result of the image processing matches the target image.

  According to a twelfth aspect of the present invention, when executed by a computer, the computer receives an image processing function for performing predetermined image processing and a correction instruction for a result image created as a result of the image processing, and the correction is performed. A target image creation function for creating a target image in which an instruction is reflected in the result image, and adjustment of parameter values related to the image processing so that a result image created as a result of the image processing matches the target image And a parameter adjustment function to achieve.

  According to a thirteenth aspect of the present invention, data processing means for performing predetermined data processing, and a correction instruction for the result data acquired as a result of the data processing are received, and target data in which the correction instruction is reflected in the result data Target data creating means for creating, and parameter adjusting means for adjusting a parameter value related to the data processing so that data acquired as a result of the data processing matches the target data.

  The invention according to claim 14 is the data processing apparatus according to claim 13, wherein the data processing means compares the reference image with an inspection target image obtained by outputting the reference image. The image processing is performed to detect the defective portion occurring in the image, and to generate image data in which the detected defective portion is displayed in a display form different from that of the non-defective portion, and obtain the result data. Such parameters define the detection sensitivity of the defective portion.

  The invention of claim 15 is the data processing apparatus according to claim 13 or claim 14, wherein the target data creating means displays the result data on a display screen, and the result is displayed on the display screen. Accepts input of correction instructions for data.

  The invention according to claim 16 is the data processing device according to any one of claims 13 to 15, wherein the parameter adjusting means obtains a result obtained under each parameter value while changing the value of the parameter. It is determined whether or not the data matches the target data, and the adjustment value of the parameter is determined to be a parameter value from which result data matching the target data is obtained.

  The invention according to claim 17 is the data processing device according to claim 16, wherein the parameter adjustment means is configured by parameter values obtained from the result data matching the target data with the adjustment values of the parameters. To the median of the set.

  The invention according to claim 18 is the data processing apparatus according to claim 16, wherein the parameter adjusting means changes the parameter adjustment value, and the result data that matches the target data changes the parameter value. To determine the parameter value when first obtained.

  The invention according to claim 19 is the data processing apparatus according to any one of claims 13 to 15, wherein the parameter adjusting means matches the target data from among a combination of one or more parameter values. A combination that provides data is extracted by a genetic algorithm, and an adjustment value of the parameter is determined as each value constituting the extracted combination.

  A twentieth aspect of the invention is the data processing device according to any one of the sixteenth to nineteenth aspects, further comprising: an adjustment range specifying unit that specifies a range in which the parameter value is changed based on an input setting value. Further prepare.

  A twenty-first aspect of the invention is the data processing device according to any one of the sixteenth to twentieth aspects, wherein a change amount specification that specifies a change amount when changing the parameter value based on an input setting value. Means.

  The invention according to claim 22 is the data processing device according to any one of claims 13 to 21, wherein the adjustment of the parameters related to the data processing is performed based on the predetermined feature amount acquired from the correction instruction. Adjustment target parameter specifying means for specifying a target parameter is provided.

  According to a twenty-third aspect of the present invention, there is provided a target data generating step of receiving a correction instruction for result data acquired as a result of data processing from an operator and generating target data in which the correction instruction is reflected in the result data; And a parameter adjustment step of adjusting a parameter value related to the data processing so that data acquired as a result of the data processing matches the target data.

  According to a twenty-fourth aspect of the invention, when executed by a computer, the computer receives a data processing function for performing predetermined data processing and a correction instruction for result data obtained as a result of the data processing, and the correction is performed. A target data creation function for creating target data in which an instruction is reflected in the result data, and a parameter value relating to the data processing is adjusted so that data acquired as a result of the data processing matches the target data And a parameter adjustment function.

  According to the first, eleventh, and twelfth aspects of the present invention, the operator can adjust the parameter by giving a correction instruction to the result image, instead of directly inputting the parameter value. That is, the parameters relating to image processing can be adjusted with an easy operation. Further, since the image value obtained as a result of the image processing is adjusted to a parameter value that matches the target image, a parameter value that provides an image processing result desired by the operator can be obtained. That is, it is possible to accurately adjust parameters relating to image processing.

  According to the invention described in claims 2 and 14, it is possible to easily and accurately adjust the parameter that defines the detection sensitivity of the defective portion occurring in the inspection target image.

  According to the third aspect of the invention, since the input of the correction instruction for the result image is received from the display screen, the operator can input the correction instruction for the result image easily and accurately. As a result, parameter adjustment can be performed more easily and accurately.

  According to the invention described in claim 4, while changing the parameter value, an appropriate parameter value (that is, a parameter value such that the result image obtained under the parameter value matches the target image) is detected. Therefore, an appropriate parameter value can be reliably detected.

  According to the fifth aspect of the present invention, the parameter adjustment value is determined to be the median value of the set composed of the parameter values from which the result image matching the target image is obtained. The value can be determined as a parameter adjustment value.

  According to the sixth aspect of the present invention, the parameter adjustment value is determined to be the parameter value when the result image that matches the target image is first obtained in the process of changing the parameter value. Can be performed.

  According to the invention described in claims 7 and 19, since the optimum combination of parameter values is extracted using the genetic algorithm, it is possible to quickly adjust the parameters.

  According to the eighth and twentieth aspects of the invention, since the range in which the parameter value is changed is specified based on the input set value, the parameter adjustment range can be narrowed down. As a result, parameter adjustment can be performed quickly.

  According to the ninth and twenty-first aspects of the present invention, since the amount of change when changing the parameter value is specified based on the input set value, it is possible to narrow down the possible parameter values in the parameter adjustment range. Become. As a result, parameter adjustment can be performed quickly.

  According to the invention described in claim 10, since the adjustment target parameter is specified based on the predetermined feature amount acquired from the correction instruction, the type of parameter to be adjusted to obtain a result image that matches the target image is determined. You can choose appropriately. As a result, parameter adjustment can be performed quickly and accurately.

  According to the invention described in claims 13, 23, and 24, the operator can adjust the parameter by giving a correction instruction to the result data, instead of directly inputting the parameter value. That is, the parameters related to data processing can be adjusted with an easy operation. Further, since the parameter value is adjusted so that the data obtained as a result of the data processing matches the target data, a parameter value that provides the data processing result desired by the operator can be obtained. That is, the parameters related to data processing can be adjusted accurately.

  According to the fifteenth aspect of the present invention, since the input of the correction instruction for the result data is accepted from the display screen, the operator can input the correction instruction for the result data easily and accurately. As a result, parameter adjustment can be performed more easily and accurately.

  According to the invention of claim 16, an appropriate parameter value (that is, a parameter value that results data obtained under the parameter value matches with the target data) is detected while changing the parameter value. Therefore, an appropriate parameter value can be reliably detected.

  According to the invention described in claim 17, since the adjustment value of the parameter is determined to be the median value of the set composed of the parameter values from which the result data matching the target data is obtained, the optimum value among the appropriate parameter values is determined. The value can be determined as a parameter adjustment value.

  According to the eighteenth aspect of the invention, the parameter adjustment value is determined to be the parameter value when the result data that matches the target data is first obtained in the process of changing the parameter value. Can be performed.

  According to the invention described in claim 22, since the adjustment target parameter is specified based on the predetermined feature amount acquired from the correction instruction, the type of parameter to be adjusted to obtain result data that matches the target data is determined. You can choose appropriately. As a result, parameter adjustment can be performed quickly and accurately.

[First Embodiment]
<1. Constitution>
FIG. 1 is a diagram showing a configuration of a printing system 100 including a defect inspection apparatus 1 which is an aspect of a data processing apparatus according to the first embodiment of the present invention. As will be described in detail later, the defect inspection apparatus 1 is an apparatus that detects a defect that has occurred in a printed matter that has been created, and has a function of adjusting image processing parameters related to the defect detection process. .

<1-1. Printing system configuration>
In the printing system 100, a defect inspection device 1, a print data creation device 2, a plate making device 3, an output device 4, and an image scanner 5 are electrically connected to each other via a network N such as a LAN (Local Area Network). It is a system connected to. In the printing system 100, a series of printing workflows from creation of print data to plate making and output is performed.

  The defect inspection apparatus 1 is an apparatus that performs plate inspection / printed object inspection for detecting defects generated in a printed material, a plate-making film, a printing plate, or the like obtained by outputting a reference image. Hereinafter, a printed material, a plate making film, a printing plate, or the like, which is an inspection target for plate inspection / printed material inspection, is collectively referred to as “inspection object”. Plate inspection / printed material inspection refers to image data of an inspection object (hereinafter referred to as “inspection data D1”) and image data used as a reference for generating the inspection object (hereinafter referred to as “reference data D0”). This is a process for detecting a difference from that as a defect.

  The print data creation device 2 is a device that creates print data. More specifically, layout data such as image layout in a print image is generated to generate layout data, and the generated layout data is converted into multi-gradation image data by RIP processing (rasterization processing). The RIP process may be performed by an independent device different from the print data creation device 2.

  The print data creation device 2 can generate image data of various resolutions according to the purpose of use from the same layout data. For example, high-resolution halftone image data of about 2400 dpi is generated as image data used for output, while multi-tone image data of coarse resolution of about 300 dpi is generated as reference data D0 used for inspection. be able to. However, it is desirable that the resolution of the reference data D0 is determined according to the inspection object data D1, and in this embodiment, the resolution of the reference data D0 is set to the same level as that of the inspection object data D1. For example, when the printed data is photographed with a digital camera having a resolution of 300 dpi and the inspection object data D1 is acquired, the print data creation device 2 creates the reference data D0 by RIP-developing the layout data at a resolution of 300 dpi. The created reference data D0 is sent to the defect inspection apparatus 1 via the network N and used for plate inspection / printed material inspection.

  The plate making apparatus 3 is an apparatus (so-called CTP apparatus) that outputs a printing plate based on halftone image data. More specifically, a printing plate is created by forming a printing image based on halftone image data on a printing plate by laser exposure. In addition, the platemaking film based on halftone image data may be created by an image setter, and the printing plate may be created using the created platemaking film. In this case, the plate making apparatus 3 includes an image setter.

  The output device 4 is a device that performs printing on printing paper using the printing plate created in the plate making device 3. Note that the digital output may be performed directly on the printing paper from the halftone image data without using the printing plate.

  The image scanner 5 is an apparatus that reads an inspection target for plate inspection / printed material inspection and acquires inspection target data D1. The acquired inspection object data D1 is sent to the defect inspection apparatus 1 via the network N and used for plate inspection / printed material inspection.

<1-2. Configuration of defect inspection system>
The defect inspection apparatus 1 is realized by a computer. More specifically, the defect inspection apparatus 1 includes a CPU 11a, a ROM 11b, and a RAM 11c. The defect inspection apparatus 1 includes a control unit 11 that realizes functions described later, a program P for causing the computer to function as the defect inspection apparatus 1, and the like. A storage unit 12 for storing, an operation unit 13 including a mouse and a keyboard for an operator to input various instructions, a display unit 14 such as a display, a hard disk, and the like are configured through a media reader / writer 151. The R / W unit 15 for reading / writing data from / to the recording medium M and the communication unit 16 serving as an interface for transferring data to / from other devices on the network N are mainly used. I have.

  In the defect inspection apparatus 1, a so-called GUI (Graphical User Interface) 101 that can perform processing while displaying the operation content through the operation unit 13 and the processing status of various processes on the display unit 14. (See FIG. 2) is realized by the functions of the control unit 11, the operation unit 13, and the display unit 14. Thus, the operator can give an instruction to the computer by performing a predetermined operation using the operation unit 13 while referring to the screen displayed on the display unit 14.

<1-3. Configuration related to parameter adjustment>
The defect inspection apparatus 1 has a function of adjusting image processing parameters related to plate inspection / printed material inspection. FIG. 2 is a diagram illustrating a configuration related to the parameter adjustment function. The defect inspection apparatus 1 includes a reference data acquisition unit 110, an inspection target data acquisition unit 120, a defect detection processing unit 130, a target data creation unit 140, an adjustment target parameter identification unit 150, and an adjustment as a configuration related to parameter adjustment. A constant determination unit 160 and a parameter adjustment processing unit 170 are provided. Each of these units is a component realized by the control unit 11 executing a predetermined program P stored in the storage unit 12.

  The reference data acquisition unit 110 acquires reference data D0 in accordance with an operator instruction received via the operation unit 13. More specifically, the reference data D0 is acquired from the print data creation device 2 via the network N. Further, as shown in FIG. 1, reference data D0 recorded on a portable recording medium M (for example, MO (magneto-optical disk) or CD-R / RW) is converted into MO drive or CD-R / RW. It is also possible to acquire the reference data D0 by reading from a media reader / writer 151 such as a drive. The reference data acquisition unit 110 stores the acquired reference data D0 in the storage unit 12.

  The inspection target data acquisition unit 120 acquires the inspection target data D <b> 1 in accordance with an operator instruction received via the operation unit 13. More specifically, the inspection object data D1 is acquired from the image scanner 5 via the network N. It is also possible to read the recording medium M on which the inspection target data D1 obtained by photographing the inspection target with a digital camera or the like is read by the media reader / writer 151 to acquire the inspection target data D1. The inspection target data acquisition unit 120 stores the acquired inspection target data D1 in the storage unit 12.

  FIG. 3A illustrates the reference data D0. Further, FIG. 3B illustrates the inspection target data D1 of the inspection target created based on the reference data D0 illustrated in FIG. 3A. As described above, the reference data D0 is image data serving as a reference for creating a printed material, and the inspection target data D1 is image data of the inspection target. Originally, it is desirable that the inspection object data D1 completely coincide with the reference data D0. However, due to various factors that may occur in the printing process (for example, dust contamination or positional deviation), as shown in FIG. 3, a portion (defect) that differs from the reference data D0 occurs in the inspection target data D1. there is a possibility. The defect detection processing unit 130 described later detects this defect by executing a defect detection algorithm described later.

  The defect detection processing unit 130 compares the reference data D0 stored in the storage unit 12 with the inspection target data D1 obtained by outputting the reference data D0, and detects a defective portion generated in the inspection target data D1. Then, image processing for creating result data D2 in which the detected defective portion is displayed in a display form different from that of the non-defective portion is performed. More specifically, a predetermined defect detection is performed on the reference data D0 and the inspection target data D1 stored in the storage unit 12 to detect defects occurring in the inspection target data D1. More specifically, the defect detection image processing is blurring processing, swaying processing, comparison processing, and isolated point removal processing. These image processing is a series of algorithms for detecting defects (hereinafter referred to as “defects”). The detection algorithm is executed in a predetermined order.

  The defect detection processing unit 130 includes a blur processing unit 131 that is a processing unit that performs defect detection image processing, a sway processing unit 132, a comparison processing unit 133, and an isolated point removal processing unit 134. In addition, a result image creation unit 135 that creates the result data D2 by reflecting the detection result is provided.

  The blur processing unit 131 executes blur processing according to a parameter (blur parameter P1) related to the blur processing. The blurring process is a process of blurring the reference data D0 and the inspection target data D1, and more specifically, an average value is calculated and calculated for a pixel to be processed and a predetermined number of pixels in the vicinity thereof. This is done by using the obtained value as the output value of the pixel. Here, the degree of blurring is defined by the pixel range (more specifically, blur radius) that is included in the calculation of the average value, and this blur radius is set as the blur parameter P1. By executing blur processing in a defect detection algorithm, both image data can be properly compared with the same degree of blur.

  The swaying processing unit 132 executes the swaying process according to the parameter related to the swaying process (swaying parameter P2). The swaying process is a process of comparing the two images by virtually shifting the arrangement position of one of the reference data D0 and the inspection target data D1, and is a range of pixels to be swayed (more specifically, Is set as a comparison parameter P2. This is a case where the arrangement position of the line drawing or the pattern in the inspection target data D1 is deviated from the arrangement position in the reference data D0 (when so-called pixel deviation occurs) by executing the sway process by incorporating it in the defect detection algorithm. However, it is possible to cancel the pixel shift and detect the original difference value.

  The comparison processing unit 133 executes the comparison process according to the parameter related to the comparison process (comparison parameter P3). The comparison process is a process for detecting a difference in gradation generated between the reference data D0 and the inspection target data D1 as a defect. More specifically, with respect to the reference data D0 and the inspection target data D1, the difference value of the gradation difference is calculated for each corresponding pixel, and if the calculated difference value is larger than a predetermined threshold value, the pixel is determined to be defective. Judge. A threshold (that is, a gradation margin) that defines a gradation difference to be detected as a defect is set as the comparison parameter P3.

  The isolated point removal processing unit 134 executes the isolated point removal process in accordance with a parameter related to the isolated point removal process (isolated point removal parameter P4). The isolated point removal process is a process for removing an isolated point as an unnecessary isolated point when a pixel having a difference between the reference data D0 and the inspection target data D1 exists in a small pixel range. It is. Here, a pixel range that is determined as an isolated point (more specifically, the number of isolated point removal pixels that defines the number of pixels that form an isolated point) is set as the isolated point removal parameter P4.

  The defect detection algorithm in this embodiment is performed by executing predetermined image processing on the reference data D0 and inspection target data D1 read by the processing units 131 to 134 from the storage unit 12. More specifically, for example, the following flow is performed.

  First, the blurring processing unit 131 performs a blurring process on the reference data D0 and the inspection target data D1. By executing the blurring process, the reference data D0, which is image data obtained by RIP development, is corrected so as to have the same degree of blur as the inspection target data D1 obtained by imaging with a camera or the like. .

  Subsequently, the swaying processing unit 132 performs swaying processing on the inspection target data D1, and the comparison processing unit 133 performs comparison processing on the inspection target data D1 and the reference data D0 on which the swaying process has been performed. Execute. Thus, a significant gradation difference is detected after canceling the pixel shift.

  Further, the isolated point removal processing unit 134 executes an isolated point removal process to remove unnecessary isolated points generated as a result of the comparison process. This avoids detecting even a small isolated point that should not be detected as a defect as a defect.

  Refer to FIG. 2 again. The result image creation unit 135 creates result data D2. The result data D2 is a non-defect portion in the inspection target data D1 (that is, a portion extracted as a defect in the inspection target data D1 as a result of executing the defect detection algorithm on the reference data D0 and the inspection target data D1). The image data is characteristically displayed in a display form different from the defective portion (that is, the portion other than the portion extracted as the defect).

  FIG. 4A illustrates result data D2 obtained by executing the defect detection algorithm on the reference data D0 and the inspection target data D1 illustrated in FIG. In the result data D2, the non-defective portion, that is, the region (appropriate regions S1, S2, S3) in which it is determined that there is no significant difference from the reference data D0 in the inspection target data D1, is displayed lightly. The areas (defect areas T1, T2) that are determined to have a significant difference with respect to the portion, that is, the reference data D0 are fattened by a fattening process and then a predetermined color (for example, magenta color). Is displayed.

  The target data creation unit 140 receives a correction instruction for the result data D2 from the operator, and creates image data (hereinafter referred to as “target data D3”) in which the content of the instruction is reflected in the result data D2.

  FIG. 4B illustrates target data D3 created by receiving a correction instruction from the operator for the result data D2 illustrated in FIG. For example, when receiving an instruction from the operator to correct the defect area T2 to the appropriate area among the defect areas T1 and T2 in the result data D2 shown in FIG. ), Target data D3 in which the defect area T2 is corrected to an appropriate area is created. More specifically, the portion of the defect area T2 in the result data D2 is corrected to a thinly displayed state as an appropriate area. The operator can give a correction instruction for the result data D2 from the screen on which the result data D2 is displayed (see FIG. 7). This will be described more specifically later.

  The adjustment target parameter specifying unit 150 specifies one or more parameters (hereinafter referred to as “adjustment target parameters”) to be adjusted from the parameters P1 to P4 related to the defect detection image processing. More specifically, the parameter to be adjusted is specified by receiving selection input of a parameter desired to be adjusted among the parameters P1 to P4 from the operator.

  The adjustment constant determination unit 160 determines the “adjustment width” and “number of divisions” of each adjustment target parameter. More specifically, the adjustment width and the number of divisions are determined by receiving input of values of the adjustment width and the number of divisions from the operator. However, the “adjustment width” is an adjustment range of the parameter to be adjusted (that is, a parameter value range changed by the parameter adjustment processing unit 170 described later), and is defined by the minimum value and the maximum value of the parameter value. The “number of divisions” is a change width of the parameter value (that is, a change amount when the parameter adjustment processing unit 170 changes the parameter value).

  For example, when the adjustment range of the blur parameter P1 is set to “maximum value = 8 pix” and “minimum value = 1 pix” and the number of divisions is set to “1 pix” (see FIG. 9), the parameter adjustment processing unit 170 While changing the value of the parameter P1 from “1 pix” to “8 pix” for each “1 pix”, the optimum value of the blurring parameter P 1 is determined. Hereinafter, a parameter value that can be taken when the adjustment range defined by the adjustment range is changed by the change range defined by the number of divisions is referred to as a “trial value”. That is, the parameter adjustment value is determined as one of the trial values.

  If “0” is set as the minimum value, the parameter adjustment value may be determined to be “0”. In this case, the adjustment process in which the parameter value is set to “0” is not substantially performed in the defect detection algorithm.

  The parameter adjustment processing unit 170 determines an adjustment value of the adjustment target parameter. The parameter adjustment processing unit 170 includes a parameter value suitability determination unit 171 and a parameter value determination unit 172.

  The parameter value suitability determination unit 171 changes the parameter value in order (more specifically, changing the parameter value within the adjustment range defined by the adjustment range with the change range defined by the number of divisions (further, In other words, in the parameter value group (P1, P2, P3, P4) composed of the values of the parameters P1 to P4 related to the defect detection image processing (that is, the combination of the values of the parameters P1 to P4), It is determined whether the result data D2 obtained under each parameter value matches the target data D3. That is, for all combinations of the trial values of the parameters P1 to P4, it is determined whether or not each combination is a combination of parameter values that gives result data D2 that matches the target data D3. However, the determination as to whether or not the target data D3 and the result data D2 match is made more specifically by determining whether or not the defective areas of both image data match. The parameter value from which the data D2 obtained as a result of determining that the parameter value suitability determining unit 171 matches the target data D3 is hereinafter referred to as “appropriate value”. In addition, the parameter value group that gives the result data D2 that matches the target data D3 is hereinafter referred to as an “appropriate parameter value group”. That is, the parameter values constituting the appropriate parameter value group are “appropriate values”. For example, when the adjustment target parameter is the parameter P3 among the parameters P1 to P4, it is assumed that the defect detection algorithm is executed under a parameter value group in which the value of the adjustment target parameter P3 is “n”. If the defect area portion of the result data D2 obtained there coincides with the defect area portion of the target data D3, it can be said that this parameter value group is an appropriate parameter value group. Further, it can be said that the parameter value “n” is an appropriate value for the parameter P3.

  The parameter value determination unit 172 determines a value (that is, an adjustment value) to be output as an adjustment result of the adjustment target parameter. More specifically, it is determined to be the median value of a set composed of appropriate parameter value groups. That is, a median value of a set composed of appropriate values for each adjustment target parameter is selected and determined as an adjustment value. For example, when there is one type of parameter to be adjusted, as shown in FIG. 11A, the trial value Tao, which is the median value of the set Q1 composed of appropriate values Ta, is selected and determined as the adjustment value.

<2. processing>
Next, parameter adjustment processing in the defect inspection apparatus 1 will be described.

<Overall flow of parameter adjustment processing>
FIG. 5 is a diagram illustrating an overall flow of parameter adjustment processing (more specifically, adjustment processing of parameters P1 to P4 related to defect detection image processing) in the defect inspection apparatus 1.

  First, the reference data acquisition unit 110 and the inspection target data acquisition unit 120 acquire the reference data D0 (see FIG. 3A) and the inspection target data D1 (see FIG. 3B), respectively, and store the storage unit 12. (Step S1).

  Subsequently, the defect detection processing unit 130 executes a defect detection algorithm on the reference data D0 and the inspection target data D1 acquired in step S1, detects a defect in the inspection target data D1, and reflects the detection result. Result data D2 (see FIG. 4A) is created (step S2).

  Subsequently, the target data creation unit 140 receives a correction instruction for the result data D2 from the operator, and creates target data D3 (see FIG. 4B) in which the content of the instruction is reflected in the result data D2 (step S3). ).

  Subsequently, the adjustment target parameter specifying unit 150 specifies the adjustment target parameter from the parameters P1 to P4 related to the defect detection image processing, and determines the adjustment width and the number of divisions of each adjustment target parameter (step S4). .

  Subsequently, the parameter adjustment processing unit 170 determines an adjustment value of the adjustment target parameter (step S5).

  Subsequently, the parameter adjustment processing unit 170 is obtained when the defect detection algorithm is executed under the adjustment value determined in step S5, the target data D3 created in step S3, and the determined parameter value. The result data D2 is displayed on the display unit 14 (step S6).

  When an instruction to set the adjustment target parameter value to the displayed adjustment value is received from the operator who has seen the display result of step S6 (YES in step S7), the parameter adjustment processing unit 170 displays the adjustment target parameter value. The value is set to the adjustment value determined in step S5 (step S8). This completes the parameter value adjustment process. Below, each process of step S3, S4, S5 is demonstrated more concretely.

<2-1. Processing for creating target data D3>
The process of step S3 will be described more specifically. FIG. 6 is a diagram showing the flow of processing in step S3.

  First, the target data creation unit 140 displays a correction instruction reception screen G for the result data D2 on the display unit 14 (step S11). FIG. 7 is a diagram illustrating a configuration example of the reception screen G. The reception screen G includes a field g1 for displaying the entire result data D2 to be corrected, a field g2 for displaying a region to be corrected in the result data D2, and a selection panel for determining correction contents. And a field g3 for displaying.

  First, the target data creation unit 140 accepts designation of an area to be corrected in the field g1 (step S12). That is, the operator looks at the entire result data D2 displayed in the field g1, determines whether there is a part to be corrected in the result data D2, and if a part to be corrected is found, corrects it by dragging the mouse. Specify the target area. The target data creation unit 140 accepts the area designation. The area designation may be performed by various pointing devices other than the mouse (for example, a trackball, a touch pad, a tablet, etc.).

  When receiving the designation of the area to be corrected in the field g1, the target data creation unit 140 subsequently enlarges and displays the designated area in the field g2 (step S13). For example, as shown in FIG. 7, when the operator designates the area U0 in the result data D2 displayed in the field g1, the target data creation unit 140 accepts the designation of the area U0 and receives the area U0 in the field g2. The inside is enlarged and displayed.

  Subsequently, the target data creation unit 140 accepts designation of an area for executing a correction instruction in the field g2 (hereinafter referred to as “correction target area U”) (step S14). That is, the operator is caused to specify the portion to be corrected in the result data D2 in the field g2 more finely by dragging the mouse or the like, and the target data creation unit 140 accepts the specified region as the correction target region U. Since the partial area of the result data D2 is enlarged and displayed in the field g2, the operator can specify an area that is finer than the area specified in the field g1.

  When the area specification of the correction area U is received, the target data creation unit 140 subsequently receives an input of the content of correction to be performed on the correction target area U (step S15). More specifically, this input operation is performed when the operator selects and clicks one of the “addition” and “deletion” panels for inputting correction contents among the selection panels displayed in the field g3. .

  When the input of the correction content is received, the target data creation unit 140 executes the correction of the instructed content with respect to the result data D2 (step S16). For example, when the operator selects the “delete” panel, the target data creation unit 140 corrects the defect area included in the correction target area U into an appropriate area. When the operator selects the “add” panel, the target data creation unit 140 corrects the appropriate area included in the correction target area U to a defective area. Hereinafter, in the correction target area U, an area that is actually changed to a defective area or an appropriate area is referred to as a “difference area V”.

  In the above description, once the designation of the correction target area U is received in the field g2 (step S14), the correction is performed on the difference area V included therein (step S16). It is also possible to accept the designation of the difference area V. In this case, for example, the operator can give an instruction to correct a part of the appropriate area or the defective area displayed in the field g2 to the defective area or the appropriate area (for example, refer to the correction instruction E3 in FIG. 13).

  Subsequently, the target data creation unit 140 displays in the field g1 the result data D2 reflecting the correction content received in step S15 (step S17). That is, the display mode of the difference area V is changed and displayed in the field g1.

  Subsequently, the target data creation unit 140 waits for an instruction input for completion of correction (step S18). More specifically, in this instruction input operation, the operator finishes the correction instruction in the selection panel displayed in the field g3, and shifts to a screen for setting details of parameters to be adjusted. This is done by selecting and clicking the panel “Details” for inputting the instruction. The operator looks at the image displayed in the field g1 to determine whether or not there is a part to be further corrected. If a part to be further corrected is found, the operator selects the field “g1” without selecting the “detail” panel. Specify the area to be corrected again. In this case, the target data creation unit 140 performs the processing of steps S12 to S17 again and receives a correction instruction for the result data D2. On the other hand, when the operator determines that the instruction input for all the correction points has been completed, the operator selects the “detail” panel and inputs an instruction to end the correction instruction.

  When the input of the end of the correction instruction is accepted (YES in step S18), the target data creation unit 140 completes the creation of the target data D3 and stores the created target data D3 in the storage unit 12 (step S19). Above, the process of step S3 is complete | finished.

<2-2. Identifying parameters to be adjusted>
The process of step S4 will be described more specifically. FIG. 8 is a diagram showing the flow of processing in step S4.

  When the process of step S3 is completed (that is, when an operation on the “details” panel is accepted on the acceptance screen G), the adjustment target parameter specifying unit 150 displays a setting input acceptance screen H for parameter adjustment on the display unit 14. (Step S21). FIG. 9 is a diagram illustrating a configuration example of the reception screen H. The reception screen H includes a field h1 for receiving an input for selecting an adjustment target parameter, a field h2 for receiving an input of an adjustment width and the number of divisions, and a field h3 for displaying a selection input / decision selection panel. .

  First, the adjustment target parameter specifying unit 150 accepts selection of a parameter to be adjusted in the field h1 (step S22). More specifically, the types of parameters P1 to P4 related to the defect detection image processing are displayed in the field h1, and check boxes for the parameters P1 to P4 are displayed. The operator can select a parameter desired to be adjusted from parameters P1 to P4 by checking a check box by operating the mouse.

  In the example of FIG. 9, the field h1 includes an item “blur radius” indicating the blur parameter P1, an item “sway radius” indicating the blur parameter P2, and “isolation point removal” indicating the isolated point removal parameter P4. “Number of pixels” is displayed. In addition, although hidden in FIG. 9, when the bar displayed on the right side of the field h1 is dragged, an item of “gradation margin” indicating the comparison parameter P3 is further displayed. The operator can select the parameter desired to be adjusted by checking the checkbox C of the desired parameter.

  When the operator checks one of the check boxes and selects a parameter, the adjustment target parameter specifying unit 150 receives input of the adjustment width and the number of divisions for the parameter selected in step S22 in the field h2 (step S23). ). More specifically, an input box for an adjustment width (that is, a maximum value and a minimum value of the adjustment range) and an input box for the number of divisions are displayed in the field h2. The operator can input a desired adjustment width and the number of divisions by inputting a value in each input box.

  For example, as shown in FIG. 9, when the operator checks the “blur radius” item and selects the blur parameter P1 as the adjustment target parameter, the adjustment target parameter specifying unit 150 displays the “blur radius” display portion. And an input box for the adjustment width of the blurring parameter P1 and an input box for the number of divisions are displayed in the field h2. FIG. 9 shows a state where the operator inputs the maximum value “8 (pix)”, the minimum value “1 (pix)” as the adjustment range of the blurring parameter P1, and “1 (pix)” as the number of divisions. Yes.

  Subsequently, the adjustment target parameter specifying unit 150 waits for input of selection of the adjustment target parameter and completion of setting input of the adjustment width and the number of divisions for each adjustment target parameter (step S24). More specifically, in this instruction input operation, the operator selects and clicks a panel “decision” for inputting an instruction to complete the detailed setting of the parameter among the selection panels displayed in the field h3. Is done by. When the operator determines that there are still parameters to be added to the adjustment target, the operator further selects parameters in the field h1 without selecting the “decision” panel. In this case, the adjustment target parameter specifying unit 150 performs the processes of steps S22 to S24 again and receives a detailed setting input for the parameter. On the other hand, when the operator determines that the detailed setting for the parameter has been completed, the operator selects the “decision” panel and inputs an instruction to end the detailed setting.

  When an instruction to end the detailed setting is received (YES in step S24), the adjustment target parameter specifying unit 150 reads the parameter checked in the check box as the adjustment target parameter (step S25), and is further set for each adjustment target parameter. The adjustment width and the number of divisions are read (step S26).

  The operator can click on the cancel panel displayed in the field h3 and redo the selection of parameters and the input of the adjustment width and the number of divisions. When the cancel panel is operated, the adjustment target parameter specifying unit 150 invalidates the operation of the previous operator. Above, the process of step S4 is complete | finished.

<2-3. Adjustment value determination process>
The process of step S5 will be described more specifically. FIG. 10 is a diagram showing the flow of processing in step S5.

  When the process of step S4 is completed (that is, when the operation of the “decision” panel is accepted on the acceptance screen H), the acceptance screen G is displayed on the display unit 14 again. When an operation on the “recalculation” panel is accepted on the acceptance screen G, the process of step S5 is started.

  First, the parameter adjustment processing unit 170 sets initial values of parameters P1 to P4 related to defect detection image processing (step S31). More specifically, among the parameters P1 to P4, the value of the adjustment target parameter is set to the minimum value of the adjustment range set for the adjustment target parameter. For parameters that are not adjustment target parameters, the currently set value is set as the initial value. The value of the parameter that is not the adjustment target is fixed at this initial value. A parameter value group constituted by the initial values of the parameters P1 to P4 is hereinafter referred to as an “initial parameter value group”.

  Subsequently, the defect detection processing unit 130 sets the set parameter value (that is, the initial parameter value group) as a determination target parameter value group, and the reference data D0 acquired in step S1 under the determination target parameter value group. The defect detection algorithm is executed on the inspection target data D1, and result data D2 reflecting the detection result is created (step S32).

  Subsequently, the result data D2 created in step S32 is stored in the storage unit 12 (step S33).

  Subsequently, the parameter value suitability determination unit 171 compares the result data D2 created in step S32 with the target data D3 created in step S3, and the defective area portion of the result data D2 is the target data D3. It is determined whether or not it coincides with the defective area (step S34). More specifically, in the result data D2, when a portion that becomes a defective area is a defective area in the target data D3 and no other defective area exists in the target data D3 (that is, the result data D2 and the target data D3 If the defect area included in the image data is completely matched), it is determined that the result data D2 and the target data D3 match. Otherwise, it is determined that the two image data do not match.

  When it is determined in step S34 that the defect area portions of the two image data match, the parameter value suitability determination unit 171 determines the parameter value (adjustment target parameter is the adjustment target parameter) employed in the defect detection algorithm executed in step S32. When there are a plurality of parameters, a combination of the parameter values employed in the defect detection algorithm is determined to be an appropriate value, and when there are a plurality of parameters to be adjusted (a combination of the parameter values). ) Is stored in the storage unit 12 (step S35). In other words, it is determined that the determination target parameter value group employed in the defect detection algorithm executed in step S32 is an appropriate parameter value group (that is, each value of the adjustment target parameter constituting the determination target parameter value group). Are determined as appropriate values), and each value of the adjustment target parameter determined as the appropriate value is stored in the storage unit 12.

  On the other hand, if it is determined in step S34 that the defect areas of the two image data do not match, the parameter value suitability determination unit 171 uses the parameter values (adjustments) employed in the defect detection algorithm executed in step S32. If there are a plurality of target parameters, it is determined that the combination of the parameter values employed in the defect detection algorithm is not an appropriate value (step S36). In other words, it is determined that the determination target parameter value group employed in the defect detection algorithm executed in step S32 is not an appropriate parameter value group.

  Subsequently, whether the value set for each adjustment target parameter (in other words, the value set for each adjustment target parameter in the determination target parameter value group) is the maximum value of the adjustment range of the parameter. Is determined (step S37).

  If it is determined in step S37 that there is at least one parameter to be adjusted that is not set to the maximum adjustment range of the parameter, the parameter value (that is, the value of the parameter constituting the parameter value group to be determined) The setting is changed (step S38). More specifically, this process is performed as follows.

  When there is one type of adjustment target parameter, the value currently set for the adjustment target parameter is changed to a value increased by the number of divisions of the adjustment target parameter, and acquired as a new determination target parameter value group To do.

  However, if the value set for the parameter to be adjusted is set to the maximum value of the adjustment range for the parameter, the suitability determination for all trial values is made when the defect detection algorithm under the parameter value is completed. It is determined that the process has been completed (that is, YES in step S37), and the process proceeds to step S39 without any further parameter setting change.

  When there are two types of adjustment target parameters, one of the adjustment target parameters (the second adjustment target parameter) remains fixed and the other adjustment target parameter (the first adjustment target parameter is used). ) Is changed to a value increased by the number of divisions of the adjustment target parameter, and acquired as a new determination target parameter value group.

  When the value set for the first adjustment target parameter is the maximum value of the adjustment range of the adjustment target parameter, the value currently set for the second adjustment target parameter is divided into the adjustment target parameter. While changing the setting to a value increased by the number, the value of the first adjustment target parameter is set again to the minimum value of the adjustment range of the adjustment target parameter, and acquired as a new determination target parameter value group.

  However, if the value set for each of the first and second adjustment target parameters is set to the maximum value of the adjustment range of the parameter, at the time when the defect detection algorithm under the parameter value is completed. Then, it is determined that the suitability determination has been completed for all combinations of adjustment values of the parameter to be adjusted (that is, YES in step S37), and the process proceeds to step S39 without any further parameter setting change. become.

  Even when there are three or more types of adjustment target parameters, the parameter values are sequentially changed until appropriateness determination is performed for all combinations of trial values of the adjustment target parameters, as in the case of two types of adjustment target parameters.

  Refer to FIG. 10 again. If it is determined in step S37 that the values set for all of the adjustment target parameters are the maximum value of the adjustment range of the parameters (that is, all combinations of trial values of the adjustment target parameters (ie, possible combinations) When the defect detection algorithm is executed for all parameter value groups), the parameter value determining unit 172 determines the adjustment value of the adjustment target parameter. More specifically, this process is performed as follows.

  First, it is determined whether there is a parameter value determined to be an appropriate value (a combination of parameter values determined to be an appropriate value when there are a plurality of adjustment target parameters) (step S39). More specifically, it is determined whether or not the parameter value determined to be an appropriate value is stored in the storage unit 12.

  If the parameter value determined to be an appropriate value is not stored in the storage unit 12, the parameter value appropriateness determination unit 171 notifies the operator that an appropriate parameter value has not been found (step S40). For example, a message “Parameter adjustment failed” is displayed on the display unit 14.

  When the parameter value determined to be an appropriate value is stored in the storage unit 12, the parameter value appropriateness determination unit 171 selects an optimum parameter value among the appropriate values (in the case where there are a plurality of adjustment target parameters, the optimum parameter value). Value combination) is selected as the adjustment value (step S41). More specifically, the median value of trial values determined as appropriate values is selected as the adjustment value.

  Specifically, this process is performed as follows. When there is only one type of adjustment target parameter, as shown in FIG. 11A, the median value Tao of the set Q1 composed of trial values Ta determined to be appropriate values among the trial values T of the adjustment target parameter. To select an adjustment value.

  In the case where there are two types of adjustment target parameters, as shown in FIG. 11B, the parameter values determined to be appropriate among the combinations T of the trial values of the first adjustment target parameter and the second adjustment target parameter. The median value Tao of the set Q2 composed of the combinations Ta is selected and determined as an adjustment value.

  When there are three or more types of adjustment target parameters, as in the case of two types of adjustment target parameters, a set of parameter values that are determined to be appropriate values among the trial value combinations of the adjustment target parameters Select the median and determine the adjustment value. Above, the process of step S5 is complete | finished.

<3. effect>
In the defect inspection apparatus 1 according to the first embodiment of the present invention, when adjusting the parameter values related to image processing, the target data creation unit 140 receives a correction instruction for the result data D2 from the operator and receives the target data D3. And the parameter adjustment processing unit 170 determines the parameter value from which the result data D2 that matches the target data D3 is obtained as the adjustment value. As a result, the operator does not need to directly set and input the parameter value (that is, without searching for and inputting the desired parameter value using a trial and error method), and simply giving a correction instruction to the result data D2. Desired parameter values can be obtained. In addition, since the parameter adjustment processing unit 170 determines the adjustment value as a value from which the result data D2 that matches the target data D3 is obtained, the parameter adjustment can be performed accurately.

  In addition, since appropriate parameter adjustment can be easily performed, parameter adjustment, which has been often performed by a serviceman task in the past, can be changed to a user task. This makes it possible to reduce maintenance costs.

  In particular, since the defect inspection apparatus 1 can easily and accurately adjust the parameters P1 to P4 that define the defect detection sensitivity, the defect inspection apparatus 1 can detect defects with the detection sensitivity desired by the operator. It becomes possible.

  In particular, since the target data creation unit 140 displays the inspection target data D1 on the display unit 14 and accepts an input of a correction instruction for the result data D2 from the display screen, the operator can easily and accurately correct the correction instruction for the result data D2. Can be entered. As a result, parameter adjustment can be performed more easily and accurately.

  In particular, the parameter value suitability determination unit 171 determines whether or not the parameter value is an appropriate value while sequentially changing the value of the adjustment target parameter (more specifically, the result obtained under each parameter value). By determining whether or not the data D2 matches the target data D3, it is determined whether or not the parameter value is an appropriate value), and the parameter value determination unit 172 sets the adjustment value of the adjustment target parameter to the appropriate value. Therefore, it is possible to reliably detect a parameter value for obtaining result data D2 that matches the target data D3.

  In particular, since the parameter value determination unit 172 selects the adjustment value of the adjustment target parameter as the adjustment value by selecting the median value of the set composed of the parameter values determined to be appropriate values, the optimum value among the appropriate values is determined. It is possible to determine the value (that is, the parameter value with which the result data D2 that matches the target data D3 most reliably is obtained) as the parameter adjustment value.

  In particular, the adjustment constant determination unit 160 defines the adjustment range of the parameter to be adjusted based on the input from the operator, so that the parameter adjustment range can be narrowed down. As a result, parameter adjustment can be performed quickly.

  In addition, since the adjustment constant determination unit 160 defines the number of divisions of the adjustment target parameter based on the input from the operator, it is possible to narrow down the parameter values that can be taken in the adjustment range. As a result, parameter adjustment can be performed more quickly.

  In particular, the adjustment target parameter specifying unit 150 specifies the parameter to be adjusted from the parameters related to the defect detection image processing based on the input from the operator, so that the type of parameter to be adjusted can be narrowed down. It becomes possible. As a result, parameter adjustment can be performed quickly.

[Second Embodiment]
<1. Constitution>
Since the defect inspection apparatus which is one aspect of the data processing apparatus according to the second embodiment of the present invention is realized in the same apparatus configuration as the defect inspection apparatus 1 described in the first embodiment, the description thereof will be given. Omitted. In the following, in showing the components of the defect inspection apparatus according to the second embodiment, the same components as those of the defect inspection apparatus 1 according to the first embodiment are the same as those in the first embodiment. The reference numerals used in the description are used.

<1-1. Configuration related to parameter adjustment>
The defect inspection apparatus according to this embodiment has a function of adjusting image processing parameters related to plate inspection / printed material inspection. FIG. 12 is a diagram illustrating a configuration related to the parameter adjustment function.

  The defect inspection apparatus according to this embodiment is different from the defect inspection apparatus 1 according to the first embodiment in that an adjustment target parameter determination unit 250 is provided instead of the adjustment target parameter specifying unit 150. In the following, only points that differ from the defect inspection apparatus 1 according to the first embodiment will be described, and descriptions of points that are not different will be omitted. In addition, the same components are denoted by the same reference numerals.

  Similar to the adjustment target parameter specifying unit 150, the adjustment target parameter determination unit 250 specifies the adjustment target parameter from the parameters P1 to P4 related to the defect detection image processing. However, the adjustment target parameter determination unit 250 does not specify the adjustment target parameter by receiving the selection input from the operator, but adjusts based on the correction instruction for the result data D2 received from the operator by the target data creation unit 140. Identify the target parameter.

  The adjustment target parameter determination unit 250 includes a feature amount acquisition unit 251, a corresponding parameter determination unit 252, and an adjustment target parameter determination unit 253.

  The feature quantity acquisition unit 251 acquires a predetermined feature quantity from each of the correction instructions (hereinafter simply referred to as “correction instructions”) for the result data D2 received from the operator by the target data creation unit 140.

  Here, the feature amount will be described more specifically with reference to FIGS. However, FIG. 13A illustrates the result data D2, and FIG. 13B receives and generates correction instructions E1 to E3 from the operator for the result data D2 illustrated in FIG. 3A. The target data D3 is illustrated.

  Here, it is assumed that the correction instruction E1 is an instruction for correcting one isolated point (difference area V1) in the correction target area U1 from the appropriate area to the defective area. The correction instruction E2 is an instruction to correct a plurality of isolated points (difference areas V2a, V2b, V2c, V2d) in the correction target area U2 from a proper area to a defective area. Further, it is assumed that the correction instruction E3 is an instruction to correct the difference area V3 from the appropriate area to the defective area.

  The feature amount of the correction instruction is a value that serves as a determination element for specifying the type of parameter to be adjusted in order to generate the result data D2 corresponding to the content of the correction instruction. The feature amount acquisition unit 251 acquires three pieces of information of area S, number A, and brightness distribution H as feature amounts from each correction instruction.

  The area S is the total area of the difference region V related to one correction instruction. For example, the area S acquired from the correction instruction E2 is the sum of the areas of the different areas V2a to V2d included in the correction target area U2. Further, for example, the area S acquired from the correction instruction E3 is the area of the difference region V3.

  The number A is the number of different areas V related to one correction instruction. For example, when there are four independent difference areas V1a to V2d in the correction target area U2 as in the correction instruction E2, the number A is “4”. On the other hand, when there is only one independent difference area V1 in the correction target area U1 as in the correction instruction E1, the number A is “1”. Further, when the difference area V is directly designated as in the correction instruction E3, the number A is “1”.

  The lightness distribution H is a lightness distribution of a predetermined region (that is, a region occupying a predetermined range from the center of the difference region V, hereinafter referred to as “lightness amount acquisition region W”) of the inspection target data D1.

  Here, acquisition of the lightness distribution H will be described more specifically. In acquiring the brightness distribution H, the adjustment target parameter determination unit 250 first specifies the brightness amount acquisition region W. More specifically, first, the center Vo of the different area V (or the center of all the different areas V when there are a plurality of different areas V related to one correction instruction) is specified, and the specified center Vo is set as the center. The brightness amount acquisition area W to be defined is defined.

  Subsequently, the adjustment target parameter determination unit 250 acquires the lightness distribution of the lightness amount acquisition region W in the inspection target data D1. For example, in the case of the correction instruction E1, the center of the brightness amount acquisition area W1 coincides with the center Vo of the difference area V1 (see FIG. 14A). Here, in the inspection target data D1, assuming that the difference area V1 is a point existing in a white background as shown in FIG. 14A, as the brightness distribution of the brightness amount acquisition area W, for example, FIG. ) Is obtained.

  Further, for example, in the case of the correction instruction E3, the center of the brightness amount acquisition area W coincides with the center Vo of the difference area V3 (see FIG. 14B). Here, when the difference area V3 is in the boundary portion between the pattern and the background as shown in FIG. 14B in the inspection object data D1, as the brightness distribution of the brightness amount acquisition area W, for example, as shown in FIG. A histogram like this is acquired.

  On the other hand, when the difference area V3 is in the area portion that changes to gradation as shown in FIG. 14C in the inspection object data D1, as the brightness distribution of the brightness amount acquisition area W, for example, as shown in FIG. A simple histogram is obtained.

  Refer to FIG. 12 again. The corresponding parameter determination unit 252 determines the type of parameter to be adjusted based on each feature amount acquired by the feature amount acquisition unit 251.

  Here, the determination method will be described more specifically. The determination based on the area S is performed as follows. First, it is determined whether or not the value of the area S is larger than a predetermined value s1. Here, when it is determined that the value of the area S is larger than the predetermined value s1, it is determined that the comparison parameter P3 should be adjusted.

  When it is determined that the value of the area S is not larger than the predetermined value s1, it is subsequently determined whether or not the value of the area S is larger than the predetermined value s2 (where s1> s2). Here, when it is determined that the value of the area S is larger than the predetermined value s2, it is determined that the blurring parameter P1 should be adjusted. On the other hand, if it is determined that the value of the area S is not greater than the predetermined value s2, it is determined that the isolated point removal parameter P4 should be adjusted.

  The determination based on the number A is performed as follows. First, it is determined whether or not the value of the number A is larger than a predetermined value n1. Here, when the number A is larger than the predetermined value n1, it is determined that the sway parameter P2 should be adjusted. On the other hand, when the number A is not larger than the predetermined value n1, it is determined that the isolated point removal parameter P4 should be adjusted.

  The determination based on the lightness distribution H is performed as follows. First, the adjustment target parameter determination unit 250 determines whether the histogram shape of the acquired brightness information is “first distribution shape h1”, “second distribution shape h2”, or “third distribution shape h3”.

  As a result, when the histogram shape is determined to be the first distribution shape h1, the isolated point removal parameter P4 is determined to be the second distribution shape h2, and when the histogram shape is determined to be the second distribution shape h2, the blurring parameter P1 is determined to be the third distribution shape h3. If so, it is determined that the comparison parameter P3 should be adjusted.

  Here, a method for determining which of the distribution shapes h1 to h3 is the histogram shape will be described with reference to FIGS. However, FIG. 16 is a diagram illustrating a flow of determining the type of histogram shape.

  First, the largest brightness value (maximum value M) in the histogram distribution is acquired (step S101).

  Subsequently, a reference value N is calculated (step S102). However, the reference value N is half the maximum value M (N = M / 2).

  Subsequently, it is determined whether or not the distribution width where the lightness value is the reference value N is larger than a predetermined value (step S103). However, as shown in FIG. 15B, when two or more independent distribution regions appear where the lightness value is the reference value N, it is determined whether or not the widest distribution width is larger than a predetermined value. .

  When it is determined in step S103 that the distribution width is larger than the predetermined value (that is, when there is a distribution width larger than the predetermined width), it is determined that the histogram distribution is the third distribution shape h3 (step S104). .

  When it is determined in step S103 that the distribution width is not larger than the predetermined value (that is, when there is no distribution width larger than the predetermined width), it is subsequently determined whether or not two or more distribution regions appear. (Step S105).

  When it is determined in step S106 that two or more distribution shapes appear, it is determined that the histogram distribution is the second distribution shape h2 (step S106).

  If it is not determined in step S106 that two or more distribution regions have appeared, the histogram distribution is determined to be the first distribution shape h1 (step S107).

  For example, when the above determination processing is performed on the histogram shapes of FIGS. 15A to 15C, it is determined that the histogram of FIG. 15A is the first distribution shape h1 and the isolated point removal parameter P4 should be adjusted. Is done. Further, the histogram of FIG. 15B is the second distribution shape h2, and it is determined that the blurring parameter P1 should be adjusted. Further, the histogram of FIG. 15C is the third distribution shape h3, and it is determined that the comparison parameter P3 should be adjusted.

  Refer to FIG. 12 again. The adjustment target parameter determination unit 253 finally identifies the adjustment target parameter by combining the determination results output by the corresponding parameter determination unit 252. More specifically, the proportion of each parameter in the accumulated determination result is calculated, and two parameters are determined as adjustment target parameters in descending order of proportion.

<2. processing>
A parameter adjustment process in the defect inspection apparatus 1 according to the second embodiment will be described. The overall flow of the parameter adjustment process is as shown in FIG. However, in this embodiment, the adjustment target parameter specifying process (that is, the process corresponding to step S4) is performed as follows.

<2-1. Identifying parameters to be adjusted>
FIG. 17 is a diagram showing the flow of the adjustment target parameter specifying process executed in this embodiment.

  First, the feature quantity acquisition unit 251 acquires a predetermined feature quantity (that is, area S, number A, brightness distribution H) from each of the correction instructions for the result data D2 received by the target data creation unit 140 in step S3. (Step S111).

  For example, when generating the target data D3, as shown in FIG. 13, when three correction instructions E1 to E3 are received, the feature amount acquisition unit 251 determines the area S and the number of pieces for each of the correction instructions E1 to E3. A, brightness distribution H is acquired.

  Subsequently, the corresponding parameter determination unit 252 determines the type of parameter to be adjusted based on each of the feature amounts acquired in step S111 (step S112). The determination results are sequentially sent from the corresponding parameter determination unit 252 to the adjustment target parameter determination unit 253, and the adjustment target parameter determination unit 253 accumulates the determination results.

  When the corresponding parameter determination unit 252 completes the determination for all the acquired feature values and determines that the adjustment target parameter determination unit 253 has acquired the determination results for all the feature values (YES in step S113), the adjustment is performed. The target parameter determination unit 253 identifies the parameters to be adjusted by integrating the accumulated determination results (step S114). More specifically, the proportion of each parameter in the accumulated determination result is calculated, and two parameters are determined as adjustment target parameters in descending order of proportion.

  For example, when the feature amount acquisition unit 251 acquires the area S, the number A, and the brightness distribution H for each of the correction instructions E1 to E3 in step S111, the corresponding parameter determination unit 252 acquires the nine feature amounts acquired in step S112. The types of parameters to be adjusted are sequentially determined based on each of the above. This determination result is sent to the adjustment target parameter determination unit 253, and the adjustment target parameter determination unit 253 accumulates nine determination results as shown in FIG. In the example of FIG. 18, the adjustment target parameter determination unit 253 adjusts the isolated point removal parameter P4, which is the parameter that occupies the largest proportion of the accumulated determination results, and the blur parameter P1 that has the next largest proportion. Determine the parameters.

  Refer to FIG. 17 again. When two adjustment target parameters are determined in step S114, the adjustment target parameter determination unit 250 displays a setting input reception screen H (see FIG. 9) for parameter adjustment on the display unit 14 (step S115). In this embodiment, the field h1 on the reception screen H functions not as a field for receiving an adjustment target parameter selection input but as a field for displaying the adjustment target parameter determination result. That is, the adjustment target parameter determination unit 250 receives the reception screen H in which the check box for the parameter determined as the adjustment target parameter in step S114 is checked among the check boxes displayed in the field h1. Is displayed. The fields h2 and h3 are the same as in the first implementation.

  Subsequently, the adjustment target parameter determination unit 250 receives input of the adjustment width and the number of divisions for each of the adjustment target parameters (step S116). This process is the same as the process of step S23 (FIG. 8) described above.

  Subsequently, when the adjustment width and the number of divisions have been input for each adjustment target parameter and the operation of the determination panel is accepted (YES in step S117), the adjustment target parameter determination unit 250 is set for each adjustment target parameter. The adjustment width and the number of divisions are read (step S118). This completes the adjustment target parameter identification processing.

<3. effect>
In the defect inspection apparatus according to the second embodiment of the present invention, in particular, the adjustment target parameter determination unit 250 becomes an adjustment target from among parameters related to defect detection image processing based on a correction instruction from the operator. Since the parameter is determined, it is possible to appropriately select the type of parameter to be adjusted in order to obtain the result data D2 that matches the target data D3. As a result, parameter adjustment can be performed quickly and accurately.

[Third Embodiment]
<1. Constitution>
The defect inspection apparatus, which is an aspect of the data processing apparatus according to the third embodiment of the present invention, is realized in the same apparatus configuration as the defect inspection apparatus 1 described in the first embodiment. Omitted. In the following description, the reference numerals used in the description of the first embodiment are used to show the components of the defect inspection apparatus according to the third embodiment.

  Similar to the defect inspection apparatus 1 according to the first embodiment, the defect inspection apparatus according to this embodiment includes an adjustment constant determination unit 170. However, in this embodiment, the adjustment constant determination unit 170 does not determine the adjustment value by selecting the median of the set composed of the appropriate values, but determines the appropriate value first detected as the adjustment value. To do.

<2. processing>
A parameter adjustment process in the defect inspection apparatus 1 according to the third embodiment will be described. The overall flow of the parameter adjustment process is as shown in FIG. However, in this embodiment, the parameter adjustment value determination process (that is, the process corresponding to step S5) is performed as follows.

<2-1. Adjustment value determination process>
FIG. 19 is a diagram showing the flow of the adjustment value determination process for the adjustment target parameter executed in this embodiment.

  First, processing similar to that described above in steps S31 to S34 (FIG. 10) is performed (steps S211 to S214).

  If it is determined in step S214 that the defect area portions of the two image data match, the parameter value suitability determination unit 171 determines the parameter value (adjustment target parameter is the adjustment target parameter) employed in the defect detection algorithm executed in step S212. In the case of a plurality of parameters, it is determined that the combination of parameter values employed in the defect detection algorithm is an appropriate value. In this case, the parameter value determined to be the appropriate value is determined as the adjustment value of the adjustment target parameter (step S215). That is, it is not necessary to execute the defect detection algorithm for all combinations of trial values of each adjustment target parameter as in the first embodiment, and when an appropriate parameter value is found, the parameter value is adjusted to the adjustment value. To decide. That is, the appropriate value detected first is determined as the adjustment value.

  On the other hand, if it is determined in step S214 that the defect areas of the two image data do not match, the parameter value suitability determination unit 171 uses the parameter values (adjustments) employed in the defect detection algorithm executed in step S212. If there are a plurality of target parameters, it is determined that the combination of parameter values employed in the defect detection algorithm is not an appropriate value (step S216).

  In this case, subsequently, it is determined whether or not the value set for each parameter to be adjusted is the maximum value of the adjustment range of the parameter (step S217).

  If it is determined in step S217 that there is at least one parameter to be adjusted that is not set to the maximum adjustment width of the parameter, the parameter value setting is changed (step S218). This process is the same as the process of step S38 (FIG. 10) described above.

  In step S217, when it is determined that the values set for all of the adjustment target parameters are the maximum value of the adjustment range of the parameter (that is, the defect detection algorithm is used for all combinations of adjustment values of the adjustment target parameters). When executed, the parameter value suitability determination unit 171 notifies the operator that an appropriate parameter value has not been found (step S219). That is, even if the defect detection algorithm is executed for all combinations of trial values of each adjustment target parameter, if an appropriate parameter value is not found, the process of step S219 is performed. This completes the adjustment value determination process for the adjustment target parameter.

<3. effect>
In the defect inspection apparatus according to the third embodiment of the present invention, the adjustment value of the parameter to be adjusted is first determined as the adjustment value that is determined to be an appropriate value, so that parameter adjustment can be performed quickly. It becomes possible.

[Fourth Embodiment]
<1. Constitution>
The defect inspection apparatus, which is an aspect of the data processing apparatus according to the fourth embodiment of the present invention, is realized in the same apparatus configuration as the defect inspection apparatus 1 described in the first embodiment. Omitted. In the following description, the reference numerals used in the description of the first embodiment are used to indicate the components of the defect inspection apparatus according to the fourth embodiment.

<1-1. Configuration related to parameter adjustment>
The defect inspection apparatus according to this embodiment has a function of adjusting image processing parameters related to plate inspection / printed material inspection. FIG. 20 is a diagram illustrating a configuration related to the parameter adjustment function.

  The defect inspection apparatus according to this embodiment is different from the defect inspection apparatus 1 according to the first embodiment in the aspect of determining parameter adjustment values. In the following, only points that differ from the defect inspection apparatus 1 according to the first embodiment will be described, and descriptions of points that are not different will be omitted. In addition, the same components are denoted by the same reference numerals.

  The defect inspection apparatus according to this embodiment has a function of adjusting image processing parameters related to plate inspection / printed material inspection. FIG. 20 is a diagram illustrating a configuration related to the parameter adjustment function. In the following, FIGS. 21 and 22 are referred to as appropriate. FIG. 21 and FIG. 22 are conceptual diagrams illustrating how adjustment values are determined using a genetic algorithm.

  The defect inspection apparatus according to this embodiment includes a parameter adjustment processing unit 470. The parameter adjustment processing unit 470 determines an adjustment value of the adjustment target parameter using a genetic algorithm. More specifically, a parameter value group that provides result data D2 that matches the target data D3 is extracted by a genetic algorithm from among a plurality of parameter value groups corresponding to all combinations of the trial values of the adjustment target parameter. To do. And the value which comprises the extracted parameter value group is determined to the adjustment value of a parameter for adjustment.

  As shown in FIG. 20, the parameter adjustment processing unit 470 includes a first generation solid group generation unit 471, an adaptive value acquisition unit 472, a pair generation unit 473, a crossover processing unit 474, a mutation execution unit 475, And an optimum parameter value determination unit 476.

  The first generation solid group generation unit 471 has a predetermined number N (where N is a predetermined natural number, and the value can be arbitrarily set. In the example of FIG. 21, N = 4). Are obtained as a first generation solid group. Each solid has a chromosome composed of M genes. However, the gene is more specifically any numerical value in the range of 0.0 to 1.0, and the numerical value of the gene possessed by each individual of the first generation is selected at random. Further, the number M of genes included in the chromosome (M = 7 in the example of FIG. 21) matches the number of parameters related to the defect detection image processing, and as shown in FIG. ,... Correspond to any of parameters relating to defect detection image processing. That is, each chromosome corresponds to a parameter value group.

  When one or more parameters selected from parameters related to defect detection image processing are used as adjustment target parameters, each gene is associated with one of the adjustment target parameters. That is, in this case, the number of genes coincides with the number of parameters to be adjusted.

  The adaptation value acquisition unit 472 acquires the adaptation value of the individual. The adaptive value acquisition unit 472 acquires the adaptive value of each of the N individuals by performing processing to be described later on each of the N solids constituting each generation. Then, the acquired adaptation value is stored in the storage unit 12 in association with each individual.

  The process for acquiring the adaptive value will be described with reference to FIG. In acquiring the adaptive value, first, each of M genes (M = 7 in the example of FIG. 22) included in the solid is interpreted, and M parameter values are acquired. That is, one parameter value group is acquired from one chromosome. The process of obtaining the parameter value by interpreting the gene is performed by executing the calculation process shown in the following (Equation 1). However, in (Equation 1), “Gi” is a gene value, and “Pi” is a parameter value associated with the gene (i = 1, 2,..., M). “Min” and “max” are the minimum value and the maximum value of the adjustment range set for the parameter, respectively. “Div” is the number of divisions set for the parameter. “Int {}” indicates a calculation process for truncating a value after the decimal point.

Pi = min + (max−min) * int {Gi * div} / div (Formula 1)
Subsequently, a defect detection algorithm is executed under the acquired parameter value group, and result data D3 is acquired. Then, the acquired result data D2 is compared with the target data D3 to calculate an adaptive value of the individual. However, this “adaptive value” is the area ratio of the portion where the result data D2 and the target image data D3 match (the ratio of the area of the matching portion to the area of the entire image). That is, the adaptation value of the individual is a value indicating the degree to which the result data D2 given by the parameter value group corresponding to the solid (chromosome) matches the target data D3, and the larger the value, the better the individual (target data D3 And the individual corresponding to the parameter value group that gives the result data D2 that is highly consistent with each other. However, the adaptive value can take a numerical value in the range of 0-1. For example, it can be said that the individual whose adaptation value is “1” is the best individual corresponding to the parameter value group that gives the result data D3 that completely matches the target data D3.

  The pair generation unit 473 generates [(N / 2) -1] sets (N is an even number. If N is an odd number, (N-1) / 2 sets) from N solids of the same generation. Generate a pair. More specifically, first, from the N solid groups of the same generation, according to the principle of natural selection using adaptive values, (N-2) (if N is an odd number, (N-1) ) Solids are selected. The individual selected here becomes a parent individual that generates the next generation individual. The principle of natural selection is the principle of preferentially handing over the genes of excellent individuals (in this case, individuals with high adaptation values) to the next generation. Among the N individuals of the same generation, This is realized by increasing the probability of selecting an individual as a parent individual. For example, the adaptive value is a selection probability of being selected as a parent individual. Moreover, it is good also as a structure which ranks an individual in order with a large adaptive value, and assigns the selection probability of a high value in an order from a high-order individual. The pair generation unit 473 pairs two different solids from (N-2) solids selected according to the above principle (or (N-1) when N is an odd number). N / 2) -1] pairs (if N is an odd number, (N-1) / 2 pairs) are generated.

  The next-generation individuals are generated by performing crossover / mutation, which will be described later, so that the number of next-generation individuals generated will be the same as the number of individuals in the previous generation (that is, Adjust the number of individuals so that the number of individuals does not change over time). In the above case, the number of individuals used for crossover (that is, the number of individuals generated by crossover as will be described later) is (N-2) when N is an even number, and when N is an odd number. Since the number is (N-1), the number of next-generation individuals to be generated is reduced as it is. Therefore, an adjustment process for replenishing this shortage is performed.

  When N is an even number, two individuals are lacking. In this case, for example, one individual is supplemented by an elite strategy (a method in which an individual with a high adaptive value is left as the next generation individual as it is), and one individual having a randomly selected gene is supplemented. On the other hand, when N is an odd number, one individual is insufficient. In this case, for example, one individual is supplemented by an elite strategy.

  Note that the number of pairs generated here may be (N / 2) sets (where N is an even number, and when N is an odd number, [(N + 1) / 2] sets). In this case, the number of individuals used for crossover is N when N is an even number, and (N + 1) when N is an odd number. Therefore, the next generation individuals generated when N is an odd number. The number will increase. Therefore, in this case, when N is an odd number, after the crossover process, one individual is randomly selected from the generated next-generation individuals and deleted to adjust the number of individuals.

  The crossover processing unit 474 crosses the paired two solids to generate a next generation solid. More specifically, as shown in FIG. 21, a gene located in the latter half of a predetermined crossover position Q (where the crossover position is a position randomly selected from the chromosome) is replaced with a new one. A solid is produced. As a result, two next-generation solids are generated from each pair. As shown in FIG. 21, there is a high possibility that the adaptation value of the individual newly generated by crossing the excellent parent individuals selected from the previous generation is even better than the adaptation value of the parent. In other words, as the generation progresses, solids with low adaptation values are naturally deceived, and there is a high possibility that excellent solids with high adaptation values will be born. However, the genes indicated by hatching in FIG. 21 mean the genes corresponding to the adaptive values. Here, a state in which an individual 2 ′ having an adaptation value “0.5” is generated in the second generation is illustrated.

  As described above, the number of individuals is adjusted so that the number of next-generation individuals generated is the same as the number of individuals of the previous generation. That is, when the number of individuals used for crossover is (N-2), the shortage of solids is determined by elite strategy and random gene selection so that the number of solids generated is N. Refill one by one. When the number of individuals used for crossover is (N-1), one shortage of solids is supplemented by an elite strategy. Furthermore, when the number of solids used for crossover is (N + 1), one individual is randomly selected from the generated next generation individuals and deleted.

  When a new population is generated by crossover, the mutation execution unit 475 causes a mutation in the gene of the generated new population with a predetermined probability. A mutation is a change in the gene value of N newly generated individuals. By mutating a gene with a predetermined probability (that is, changing the value of a gene randomly with a predetermined probability), the bias of the gene value is eliminated, and the risk of falling into a local minimum in the optimization process is avoided. Can do.

  The optimum parameter value determination unit 476 determines an adjustment value (value to be output as an adjustment result) of the adjustment target parameter. More specifically, when an individual of a predetermined generation (for example, T-th generation (where T is a natural number of 2 or more arbitrarily set)) is generated, the highest adaptive value among the individuals of the generation is obtained. The individual that has it is extracted as the optimum individual. Then, the parameter value group corresponding to the extracted chromosome of the optimum individual is determined as the optimum parameter value group, and the M parameter values constituting the optimum parameter value group are determined as the adjustment values of each parameter. In other words, M parameter values obtained by interpreting M genes included in the chromosome of the optimum individual are determined as adjustment values for each parameter. In addition, it is good also as a structure which extracts the said individual | organism | solid as an optimal individual | organism | solid when the individual | organism | solid which has an adaptation value higher than a predetermined value was born irrespective of the number of generations.

<2. processing>
A parameter adjustment process in the defect inspection apparatus according to the fourth embodiment will be described. The overall flow of the parameter adjustment process is as shown in FIG. However, in this embodiment, the adjustment value determination process (that is, the process corresponding to step S5) is performed as follows.

<2-1. Adjustment value determination process>
FIG. 23 is a diagram showing a flow of adjustment value determination processing executed in this embodiment. In the following, FIGS. 21 and 22 will be referred to as appropriate.

  When the process of step S4 (FIG. 5) is completed, the acceptance screen G is displayed on the display unit 14 again. When an operation on the “recalculation” panel is accepted on the acceptance screen G, the process of step S5 is started.

  First, the first generation solid group generation unit 471 generates N individuals (in the example of FIG. 21, individuals 1, 2, 3, and 4) constituting the first generation (step S311). Each individual generated here has a chromosome composed of M genes selected at random as described above.

  Subsequently, the adaptive value acquisition unit 472 acquires the adaptive value of each of the N individuals constituting the first generation, and stores the acquired adaptive value in association with each individual in the storage unit 12 (step S312). .

  Subsequently, the pair generation unit 473 generates [(N / 2) -1] pairs (if N is an odd number, (N-1) / 2 pairs) from the first generation N individuals. (Step S313). More specifically, from the first generation N solid groups, according to the principle of natural selection using adaptive values, (N-2) (if N is an odd number, (N-1)) ) Solids, and from the selected (N−2) (or (N−1) when N is an odd number) solids, two different solids are paired, and [(N / 2) − 1] Pairs (if N is an odd number, (N-1) / 2 pairs) are generated.

  Subsequently, for each of the pairs generated in step S313, the crossover processing unit 474 crosses the paired two solids to generate a new solid (step S314). More specifically, as shown in FIG. 21, the genes in the latter half of the predetermined crossover position randomly selected from the chromosome are replaced, and (N-2) (if N is an odd number ( N-1)) individuals are newly generated. Furthermore, the number of individuals is adjusted so that the number of next-generation individuals generated is the same as the number of individuals of the previous generation. That is, when the number of generated individuals is (N-2), one individual is supplemented by an elite strategy and one individual having a randomly selected gene is supplemented. On the other hand, when the number of generated individuals is (N-1), one individual is supplemented by an elite strategy.

  Subsequently, the mutation execution unit 475 causes a gene mutation at a predetermined probability in the new object generated in step S314 (step S315).

  N individuals are newly generated by the processing from step S314 to step S315, and the N individuals generated here (in the example of FIG. 21, individuals 1 ′, 2 ′, 3 ′, 4 ′) are generated. The solid group constituting the second generation.

  Subsequently, the optimum parameter value determining unit 476 determines whether or not a solid of a predetermined generation (T generation (where T is a natural number set arbitrarily)) has been generated (step S316). In other words, it is determined whether or not the processing from step S312 to step S315 has been executed a predetermined number of times (T-1 times).

  If it is determined in step S316 that no T-th generation solid has been generated, the process returns to step S312 again. For example, in the case of T = 3, if the second generation individual group is generated in the process of step S315, the process returns to the process of step S312 again. In this case, the individuals constituting the third generation are generated by executing the processes of steps S312 to S315 on the second-generation individuals.

  If it is determined in step S316 that a T-th generation solid has been generated, the adaptive value acquisition unit 472 acquires the adaptive values of the N individuals that constitute the T-th generation, and the optimal parameter value determination unit 476. However, the individual having the highest adaptation value among the T-th generation individuals is extracted as the optimum individual (step S317). For example, in the case of T = 3, assuming that the third generation individual group is generated in the process of step S315, the optimum parameter value determination unit 476 selects the individual having the highest adaptive value among the third generation individuals. Extract as the optimal individual.

  Note that, instead of the determination process in step S316, whether or not there is an individual having an adaptation value higher than the predetermined value after the process in step S312 (that is, after the adaptation value of each individual is acquired). Processing for determining may be executed. In this case, when it is determined that such a solid exists, the individual is extracted as the optimum individual.

  Subsequently, the optimum parameter value determining unit 476 obtains M parameter values by interpreting the M genes included in the chromosome of the optimum individual extracted in step S317, and uses the obtained values for each parameter. The adjustment value is determined (step S318). Thus, the process of step S5 (FIG. 5) is completed.

  In the above, the next-generation individuals are configured to create a pair from individuals selected based on the principle of natural selection using adaptive values and cross the pair. The method is not limited to this. For example, a method (elite strategy) of leaving an individual with a high adaptive value as a next generation individual as it is may be adopted. Further, a method of randomly generating a next generation individual may be adopted.

<3. effect>
In the defect inspection apparatus according to the fourth embodiment of the present invention, the adjustment value of the parameter to be adjusted is obtained using the genetic algorithm, so that the optimum parameter value can be obtained quickly.

  Further, in the defect inspection apparatus according to this embodiment, various parameters are associated with each of a plurality of genes included in a chromosome (that is, each individual is associated with a combination of parameter values) and adapted. An excellent individual having a high value (that is, an optimal combination of adjustment values that gives the target data D3) is obtained using a genetic algorithm. Therefore, an optimum adjustment value can be obtained quickly even if the types of adjustment target parameters increase.

[Modification]
<First Modification>
In each of the above embodiments, in the defect detection algorithm, the blurring process, the swaying process, the comparison process, and the isolated point removal process are executed in a predetermined order. However, the defect detection algorithm executed in the defect inspection apparatus 1 Is not limited to this mode. For example, in place of or in combination with each of the above image processes, a kernel designated blurring process or a fattening process can be incorporated into the defect detection algorithm. And the parameter required in the various processes incorporated can be adjusted with said aspect. That is, in the above embodiment, the blur parameter P1, the sway parameter P2, the comparison parameter P3, and the isolated point removal parameter P4 are cited as the plurality of parameters related to the defect detection image processing. Such parameters are not limited to this.

  For example, a filter radius value in the filter process, a color range to be extracted in the extraction process of a specific color range (for example, each range of L value, a value, and b value in the Lab color space), a fattening width in the fattening process, Thinning width in thinning processing, filter radius value and filter ID value in edge component detection processing, flag value specifying whether various processing such as mask composition processing and comparison processing is executed, area determination processing Various parameters such as the area range (each value of the minimum area value and the maximum area value), the value of the maximum contour length in the feature amount extraction processing, and the like can be used as the adjustment target parameters.

  In addition to parameters related to defect detection image processing, various parameters (for example, illumination angle of the image scanner 5, light source color, optical magnification, image reading, noise removal, and the like) when acquiring the inspection target data D1. Such various adjustment variables) may be included in the adjustment target parameter.

  Further, whether or not to execute the defect detection algorithm processing (ON / OFF of the algorithm processing function itself) can be included as one of the adjustment target parameters.

<Second Modification>
In each of the above-described embodiments, the case where the defect inspection apparatus 1 which is an aspect of the data processing apparatus adjusts the parameters related to the defect detection image processing has been described. Apply the above aspect to a device that performs image processing (for example, image deformation processing, noise removal processing, image conversion processing, contrast adjustment processing, etc.), and adjust parameter values related to image processing executed by the device can do. As an example, a case will be described in which the above aspect is applied to an apparatus (unevenness inspection apparatus) for inspecting unevenness occurring in a liquid crystal pattern. However, the “unevenness” of the liquid crystal pattern occurs, for example, in each color of R (red), G (green), and B (blue purple) that should have a uniform density over the entire surface in a color filter that is a liquid crystal display unit. This refers to a non-uniform part of the density.

  In the unevenness inspection apparatus, a significant unevenness portion occurring in an inspection object such as a color filter is specified by executing a predetermined unevenness detection process. However, the unevenness detection process is executed as follows, for example. First, image data of the inspection object is analyzed to extract unevenness generated in the inspection object (for example, this processing includes concentration analysis, filter processing, contrast enhancement processing, and fast Fourier transform (FFT)). And a non-uniformity extraction algorithm incorporating a binarization process or the like is performed). Then, the extracted uneven portion is evaluated to detect a significant uneven portion (for example, this process is a thickening process, a thinning process, a comparison determination determination with theoretical values such as an aspect ratio, area, and position of an image) This is performed by executing an unevenness determination algorithm in which processing and the like are incorporated). Here, the unevenness extraction accuracy and the unevenness detection sensitivity are defined by parameters relating to each process. Therefore, in order to extract unevenness with appropriate accuracy and detect unevenness with appropriate sensitivity (that is, as desired by the user), it is necessary to appropriately adjust the parameter values for processing.

  The unevenness inspection apparatus according to this modification includes a processing unit that executes unevenness detection processing as a functional unit corresponding to the defect detection processing unit 130 (FIG. 2). The processing unit detects predetermined unevenness in the image data by executing predetermined unevenness detection processing on the image data of the inspection object, and displays the detected unevenness in a characteristic manner. Image processing for creating data (that is, result data D2) is performed.

  Other functional units related to parameter adjustment are the same as those in the above-described embodiment (see FIG. 2), and the target data creation unit 140 accepts a correction instruction for the result data D2 from the operator and uses the instruction content as result data. Image data (target data D3) reflected in D2 is created, and the parameter adjustment processing unit 170 determines an adjustment value of the adjustment target parameter. Thereby, it is possible to adjust a parameter related to the unevenness detection processing.

<Third Modification>
In each of the above-described embodiments, the case where the parameter adjustment related to image processing is performed has been described. However, the parameter adjustment related to various processing on data other than image data (for example, vector data) can also be performed. . As an example of adjusting parameters related to processing for vector data, a case will be described in which the above-described aspect is applied to an operation of acquiring data for accurately specifying an object to be measured from measurement data including noise by trial and error. .

  For example, it is assumed that the measurement data is substrate thickness measurement data (data obtained by measuring the thickness of a substrate having a groove (pattern) formed on the surface thereof with a spectroscopic film thickness meter). Normally, measurement data includes noise. This film thickness measurement data also includes the influence of noise, and it is difficult to directly read the actual groove width and depth from the measurement data. Therefore, an operation for estimating the width and depth of the groove is performed by a technique using the following simulation.

  First, vector data defining an appropriate width and depth (that is, vector data expressing a groove shape) is generated, and spectral simulation is performed on the vector data. That is, measurement data obtained when the groove represented by the vector data is measured by a spectroscopic film thickness meter is virtually generated (hereinafter referred to as “virtual data”). If the virtual data matches the state obtained by removing noise from the measurement data, it can be said that the vector data giving the virtual data is data that correctly represents the width and depth of the groove. Therefore, the operator searches for vector data that gives such virtual data by trial and error (more specifically, after the obtained virtual data is compared with the measured data, it is generated between the two data. Change the shape of the vector data so as to eliminate the difference, and generate virtual data from the changed vector data again, and compare the obtained virtual data with the measurement data and search for it repeatedly ).

  The parameter adjustment function described above can be applied to this work. In this modification, a processing unit (vector data generation processing unit) that executes vector data generation processing is provided as a functional unit corresponding to the defect detection processing unit 130 (FIG. 2). The processing unit generates vector data representing a groove having a predetermined width and depth. However, the width and depth of the expressed groove are defined by the parameter values given to the vector data generation processing unit. In this modification, the parameters are used as adjustment target parameters. Virtual data is acquired by performing a predetermined spectral simulation on the vector data generated by the vector data generation processing unit. In this modification, the virtual data obtained here is the result data D2.

  Other functional units related to parameter adjustment are the same as those in the above embodiment (see FIG. 2). That is, the target data creation unit 140 receives a correction instruction for the result data D2 from the operator and creates data (target data D3) in which the content of the instruction is reflected in the result data D2. In this modification, the operator modifies the result data D2 so that the created target data D3 matches the state obtained by removing noise from the measurement data.

  Then, the parameter adjustment processing unit 170 acquires the value of the parameter that can give the result data D2 that matches the target data D3 as the adjustment value. That is, the parameter adjustment processing unit 170 acquires parameter values that can generate appropriate vector data (that is, vector data that can provide virtual data that matches the state obtained by removing noise from measurement data). . The vector data generated using this parameter value is data that correctly represents the width and depth of the groove. As described above, by using the parameter adjustment function, it is possible to easily and accurately perform an operation of estimating the actual groove width and depth from the measurement data.

<Fourth modification>
Each of the above embodiments is a technique that is effective in the field where simulation is used, as described as the third modification. For example, in various article manufacturing processes, there is a case in which a shape after processing is generated from a shape of a material before processing by simulation to check whether the shape after processing becomes a target shape. .

  The parameter adjustment function described above can be applied to this work. In this modification, a processing unit (material shape generation processing unit) that executes generation processing of shape data indicating a material shape before processing is provided as a functional unit corresponding to the defect detection processing unit 130 (FIG. 2). The processing unit generates shape data representing a predetermined material shape. However, the shape to be expressed is defined by the value of the parameter given to the material shape generation processing unit. In this modification, the parameter is set as the adjustment target parameter. In particular, the coarse / dense state of the mesh expressing the shape can be one of the adjustment target parameters. By performing simulation of a predetermined processing operation on the shape data generated by the material shape generation processing unit, shape data indicating the shape after processing the material shape is acquired. In this modification, the data obtained here is the result data D2.

  Other functional units related to parameter adjustment are the same as those in the above embodiment (see FIG. 2). That is, the target data creation unit 140 receives a correction instruction for the result data D2 from the operator and creates data (target data D3) in which the content of the instruction is reflected in the result data D2. In this modified example, the operator corrects the result data D2 so that the target data D3 to be created indicates the shape after processing desired by the operator.

  Then, the parameter adjustment processing unit 170 acquires the value of the parameter that can give the result data D2 that matches the target data D3 as the adjustment value. That is, a parameter value capable of generating data indicating an appropriate material shape (that is, a material shape capable of giving a processing shape desired by the operator) is acquired. In this way, by using the parameter adjustment function, the material shape for obtaining a desired machining shape can be estimated easily and accurately.

<Fifth Modification>
A configuration (parameter evaluation unit (not shown)) that evaluates the acquired adjustment value may be further provided as a function unit that adjusts the parameters described in the above embodiments.

  The parameter evaluation unit evaluates whether the adjustment value of the adjustment target parameter determined by the parameter adjustment processing unit 170 (FIG. 2) is an appropriate value. For example, by comparing the data (final result data) obtained by executing the defect detection algorithm under the determined adjustment value and the target data D3, the area ratio of the matching portion of both data is calculated, and the value is calculated. Acquired as the evaluation value of the adjustment value. That is, it can be said that the higher the evaluation value, the higher the degree of matching of the data obtained under the obtained adjustment value with the target data D3. The obtained evaluation value may be displayed on the display unit 14, for example, so as to make the operator know. Further, when the evaluation value is lower than the predetermined value, a configuration for finally finely adjusting the parameter value may be further provided, or the parameter adjustment processing unit 170 may execute the adjustment value calculation process again. .

<Other variations>
In each of the above embodiments, in the inspection target data D1, the defect areas T1 and T2 are displayed in magenta color, thereby adopting a mode in which the defect area is characteristically displayed. The mode of displaying the area characteristically is not limited to this. For example, an aspect in which the defective area is displayed darker than the appropriate area, an aspect in which the defective area is blinked, an aspect in which the appropriate area is displayed in a predetermined color (for example, light gray), or the like may be employed.

  Further, in each of the above-described embodiments, only when the defective area portion of the result data D2 completely matches the defective area portion of the target data D3, the result data D2 is used for creating the result data D2. Although it has been determined that the trial value is an appropriate value for the parameter to be adjusted, the degree of mismatch is predetermined even if the defective area portion of the result data D2 and the defective area portion of the target data D3 do not completely match. If it is within the permissible range, it may be determined that the two data coincide with each other, and the trial value employed in creating the result data D2 is an appropriate value for the adjustment target parameter. For example, when the area of the defect region that does not match in both data is smaller than a predetermined value, it may be determined that both data match and the trial value is an appropriate value for the adjustment target parameter.

  In each of the above embodiments, when the parameter adjustment processing unit 170 changes the parameter value in the adjustment value determination process, the adjustment range defined by the adjustment width is increased in order from the minimum value. Although it was configured (step S31 in FIG. 10), the mode of changing the parameter value is not limited to this. For example, it may be decreased in order from the maximum value of the adjustment range, or may be changed while increasing or decreasing from the median value of the adjustment range by the number of divisions. Further, a value set before adjustment may be changed as an initial value.

  In each of the first embodiments, the parameter value suitability determination unit 171 selects the median value of trial values determined to be appropriate values as adjustment values (step S41 in FIG. 10). The mode for determining the median value is not limited to this. For example, trial values determined to be appropriate values (a combination of parameter values determined to be appropriate values when there are a plurality of adjustment target parameters) are displayed in a list on the display unit 14 and any appropriate value is selected by the operator. You may let them.

  In the second embodiment, three pieces of information (area S, number A, brightness distribution H) are acquired as feature amounts. However, at least one of these feature amounts is acquired as a feature amount. Also good. Moreover, the information acquired as a feature-value is not restricted to this. For example, a drag region when the operator designates the correction target region U may be acquired as a feature amount. In addition, when the adjustment target parameter determination unit 253 identifies the parameters to be adjusted by integrating the accumulated determination results, the determination results obtained from specific feature amounts are weighted and the proportion of each parameter May be calculated.

  In the second embodiment, the adjustment target parameter determination unit 250 is provided in place of the adjustment target parameter specifying unit 150 provided in the defect inspection apparatus 1 according to the first embodiment. It is good also as a structure provided with the object parameter specific | specification part 150 and the adjustment object parameter determination part 250 collectively. In this case, the configuration may be such that the operator can select whether the adjustment target parameter is specified automatically by the adjustment target parameter determination unit 250 or whether the adjustment target parameter specification unit 150 receives the operator's selection input.

  In the second embodiment, the adjustment target parameter determination unit 253 determines the two parameters as the adjustment target parameters in order from the parameter with the highest proportion in the determination result. One or two or more parameters may be selected and determined as adjustment target parameters.

1 is a diagram illustrating a configuration of a printing system 100 including a defect inspection apparatus 1 that is an aspect of an image processing apparatus according to a first embodiment of the present invention. It is a figure explaining the structure regarding a parameter adjustment function. It is a figure which illustrates standard data D0 and inspection object data D1. It is a figure which illustrates result data D2 and target data D3. It is a figure which shows the whole flow of the parameter adjustment process in the defect inspection apparatus. It is a figure which shows the flow of the production process of the target data D3. It is a figure which shows the structural example of the reception screen G. FIG. It is a figure which shows the flow of the specific process of a parameter for adjustment. 6 is a diagram illustrating a configuration example of a reception screen H. FIG. It is a figure which shows the flow of the determination process of a parameter value. It is a figure for demonstrating selecting the median value of the trial value judged to be an appropriate value as an adjustment value. It is a figure explaining the structure regarding a parameter adjustment function. It is a figure which illustrates result data D2 and target data D3. It is a figure which illustrates the lightness amount acquisition area. It is a figure which illustrates the histogram shape of brightness distribution. It is a figure which shows the flow which determines the kind of histogram shape. It is a figure which shows the flow of the specific process of a parameter for adjustment. It is a figure which illustrates the determination result accumulate | stored in the adjustment object parameter determination part 253. FIG. It is a figure which shows the flow of the determination process of a parameter value. It is a figure explaining the structure regarding a parameter adjustment function. It is a conceptual diagram explaining the determination mode of the adjustment value using a genetic algorithm. It is a conceptual diagram explaining the determination mode of the adjustment value using a genetic algorithm. It is a figure which shows the flow of the determination process of an adjustment value.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Defect inspection apparatus 130 Defect detection process part 140 Target data preparation part 150 Adjustment object parameter specific | specification part 160 Adjustment constant determination part 170 Parameter adjustment process part 250 Adjustment object parameter determination part D0 Reference | standard data D1 Inspection object data D2 Result data D3 Target data P program

Claims (24)

Image processing means for performing predetermined image processing;
A target image creating means for accepting a correction instruction for a result image created as a result of the image processing, and creating a target image in which the correction instruction is reflected in the result image;
Parameter adjusting means for adjusting a value of a parameter related to the image processing so that a result image created as a result of the image processing matches the target image;
An image processing apparatus comprising: The image processing apparatus according to claim 1,
The image processing means detects a defective portion occurring in the inspection target image by comparing a reference image and an inspection target image obtained by outputting the reference image, and detects the detected defective portion as a non-defective portion. Perform image processing to create a result image displayed in a different display form,
The image processing apparatus, wherein the parameter relating to the image processing defines a detection sensitivity of the defective portion. The image processing apparatus according to claim 1 or 2,
The image processing apparatus, wherein the target image creating means displays the result image on a display screen and receives an input of a correction instruction for the result image from the display screen. The image processing apparatus according to any one of claims 1 to 3,
The parameter adjustment means determines whether the result image obtained under each parameter value matches the target image while changing the parameter value, and sets the parameter adjustment value as the target value. An image processing apparatus, wherein a parameter value obtained as a result image that matches an image is determined. The image processing apparatus according to claim 4,
The image processing apparatus, wherein the parameter adjustment unit determines an adjustment value of the parameter as a median value of a set composed of parameter values from which a result image matching the target image is obtained. The image processing apparatus according to claim 4,
The parameter adjustment means determines the adjustment value of the parameter as a parameter value when a result image that matches the target image is first obtained in the process of changing the value of the parameter. apparatus. The image processing apparatus according to any one of claims 1 to 3,
The parameter adjusting means extracts a combination that gives a result image that matches the target image from a combination of one or more parameter values using a genetic algorithm, and sets the parameter adjustment value as the extracted combination An image processing apparatus characterized in that the values are determined for respective values. The image processing apparatus according to any one of claims 4 to 7,
An adjustment range specifying means for specifying a range for changing the parameter value based on the input set value;
An image processing apparatus further comprising: The image processing apparatus according to any one of claims 4 to 8,
A change amount specifying means for specifying a change amount when changing the parameter value based on the input set value;
An image processing apparatus further comprising: The image processing apparatus according to claim 1,
An adjustment target parameter specifying means for specifying a parameter to be adjusted from parameters related to the image processing based on a predetermined feature amount acquired from the correction instruction;
An image processing apparatus comprising: A target image creating step of receiving a correction instruction for a result image created as a result of image processing from an operator and creating a target image in which the correction instruction is reflected in the result image;
A parameter adjustment step of adjusting a value of a parameter related to the image processing so that a result image created as a result of the image processing matches the target image;
A parameter adjustment method characterized by comprising: By being executed by a computer, the computer
An image processing function for performing predetermined image processing;
A target image creation function for accepting a correction instruction for a result image created as a result of the image processing and creating a target image in which the correction instruction is reflected in the result image;
A parameter adjustment function for adjusting a value of a parameter related to the image processing so that a result image created as a result of the image processing matches the target image;
A program that can realize Data processing means for performing predetermined data processing;
Target data creation means for accepting a correction instruction for the result data acquired as a result of the data processing, and creating target data in which the correction instruction is reflected in the result data;
Parameter adjusting means for adjusting a parameter value related to the data processing so that data acquired as a result of the data processing matches the target data;
A data processing apparatus comprising: 14. A data processing apparatus according to claim 13, comprising:
The data processing means detects a defective portion occurring in the inspection target image by comparing a reference image and an inspection target image obtained by outputting the reference image, and detects the detected defective portion as a non-defective portion. Image processing is performed to create image data displayed in a display form different from that obtained as the result data,
The data processing apparatus, wherein the parameter relating to the data processing defines a detection sensitivity of the defective portion. The data processing device according to claim 13 or 14,
The data processing apparatus, wherein the target data creating means displays the result data on a display screen and receives an input of a correction instruction for the result data from the display screen. The data processing device according to any one of claims 13 to 15,
The parameter adjustment means determines whether the result data obtained under each parameter value matches the target data while changing the parameter value, and sets the parameter adjustment value as the target value. A data processing apparatus characterized by determining a parameter value from which result data matching with data is obtained. The data processing apparatus according to claim 16, comprising:
The data processing apparatus, wherein the parameter adjustment means determines an adjustment value of the parameter as a median value of a set composed of parameter values from which result data matching the target data is obtained. The data processing apparatus according to claim 16, comprising:
Data processing characterized in that the parameter adjustment means determines an adjustment value of the parameter as a parameter value when the result data matching the target data is first obtained in the process of changing the value of the parameter apparatus. The data processing device according to any one of claims 13 to 15,
The parameter adjusting means extracts a combination that gives result data that matches the target data from a combination of one or more parameter values by a genetic algorithm, and sets the parameter adjustment value as the extracted combination A data processing apparatus characterized in that the value is determined for each of the values. A data processing device according to any of claims 16 to 19, wherein
An adjustment range specifying means for specifying a range for changing the parameter value based on the input set value;
A data processing apparatus further comprising: A data processing device according to any of claims 16 to 20, wherein
A change amount specifying means for specifying a change amount when changing the parameter value based on the input set value;
A data processing apparatus further comprising: A data processing apparatus according to any of claims 13 to 21,
Adjustment target parameter specifying means for specifying a parameter to be adjusted from parameters related to the data processing based on a predetermined feature amount acquired from the correction instruction;
A data processing apparatus comprising: A target data creation step of receiving a correction instruction for the result data acquired as a result of the data processing from an operator, and creating target data in which the correction instruction is reflected in the result data;
A parameter adjustment step of adjusting a value of a parameter related to the data processing so that data acquired as a result of the data processing matches the target data;
A parameter adjustment method characterized by comprising: By being executed by a computer, the computer
A data processing function for performing predetermined data processing;
A target data creation function for accepting a correction instruction for result data acquired as a result of the data processing, and creating target data in which the correction instruction is reflected in the result data;
A parameter adjustment function for adjusting a value of a parameter related to the data processing so that data acquired as a result of the data processing matches the target data;
A program that can realize

Images