DISPLAY CONTROL INFORMATION GENERATION DEVICE, VISUAL INSPECTION DEVICE, SUBSTRATE INSPECTION SYSTEM, DISPLAY CONTROL INFORMATION GENERATION METHOD, AND COMPUTER-READABLE PROGRAM

Abstract

To shorten the time required for verification work. [Solution] A control information generating unit of a display control information generating device generates display control information for multiple defect candidates in an inspection area based on defect candidate information. Each of the multiple defect candidate data elements in the defect candidate information includes a defect candidate position and a matching rate. The defect candidate position indicates the position on the substrate 9 of a defect candidate detected by a secondary inspection using a trained model created by machine learning from tentative defect candidates detected by a rule-based primary inspection of the substrate 9. The matching rate indicates the degree to which the defect candidate matches a predetermined defect type. The display control information includes control information for hiding the multiple defect candidates in the inspection area on the display if the highest matching rate among the multiple defect candidate data elements related to the inspection area is equal to or greater than a predetermined matching rate threshold. This reduces the time required for the verification process.

Description

[Technical Field]

[0001] The present invention relates to a technique for generating display control information for displaying defect candidates on a substrate on a display device. [Background technology]

[0002] In conventional printed wiring board inspections, an inspection device inspects images of each area of ​​the printed wiring board and detects defect candidate images that are determined to have defects. An operator then visually checks the areas on the printed wiring board that correspond to the defect candidate images (a defect confirmation process known as a verify process). The defect candidates detected by the inspection device include both fatal true defects that are likely to cause quality problems and false reports (i.e., non-fatal normal defects) that are virtually unlikely to cause quality problems. In the defect confirmation process, an operator visually detects true defects from the defect candidates detected by the inspection device.

[0003] Furthermore, Patent Document 1 proposes a visual inspection system that includes a primary inspection unit that performs defect judgment based on an image of an object without using machine learning, and a secondary inspection unit that uses a machine learning model to distinguish between true defective products (true defects) and over-judged products (false reports). This prevents over-judged products from being detected as defective in the visual inspection, thereby improving productivity. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent Publication No. 2021-177154 Summary of the Invention [Problem to be solved by the invention]

[0005] The above-mentioned verification process for printed wiring boards is usually performed on a panel with one or more sheets attached, each of which has multiple pieces that will become printed wiring boards attached to it. Pieces that are found to have fatal true defects during the verification process are discarded.

[0006] When the inspection device detects multiple defect candidate images for a single piece, the operator sequentially checks the multiple defect candidate images and, when a true defect is detected, decides to discard the piece. As a result, the time spent checking other defect candidate images for the single piece before checking the defect candidate image containing the true defect is wasted.

[0007] The present invention has been made in view of the above-mentioned problems, and has as its object to shorten the time required for the verify operation. [Means for solving the problem]

[0008] A first aspect of the present invention is a display control information generating device that generates display control information for displaying defect candidates on a substrate on a display device, the display control information generating device comprising: a storage unit that stores defect candidate information including a plurality of defect candidate data elements corresponding to a plurality of defect candidates in an inspection area on the substrate; and a control information generating unit that generates display control information for the plurality of defect candidates in the inspection area based on the defect candidate information. Each of the plurality of defect candidate data elements includes a defect candidate position indicating the position on the substrate of a defect candidate detected by a second inspection using a trained model created by machine learning from tentative defect candidates detected by a rule-based first inspection of the substrate, and a matching rate indicating the degree to which the defect candidate matches a predetermined defect type. The display control information includes control information that, if the highest matching rate among the plurality of defect candidate data elements related to the inspection area is equal to or greater than a predetermined matching rate threshold, hides the plurality of defect candidates in the inspection area on the display device.

[0009] A second aspect of the present invention relates to the display control information generating device of the first aspect, wherein an inspection area group is set on the substrate, the inspection area group being a plurality of inspection areas including the inspection area. The defect candidate information includes a plurality of defect candidate data elements corresponding to a plurality of defect candidates in each inspection area of ​​the inspection area group, and each of the defect candidate data elements including the defect candidate positions and the matching rates. The display control information includes control information for hiding the plurality of defect candidates in each inspection area on the display device when the highest matching rate among the plurality of defect candidate data elements for each inspection area is equal to or greater than the matching rate threshold. The display control information includes control information for hiding all defect candidates included in the inspection area group on the display device when the plurality of defect candidates are hidden in a predetermined number or more of the plurality of inspection areas.

[0010] A third aspect of the present invention is the display control information generating device of the first aspect (which may be either the first or second aspect), wherein each of the plurality of defect candidate data elements further includes an other matching rate indicating the degree to which the one defect candidate matches another defect type different from the defect type. The display control information includes control information for hiding the plurality of defect candidates in the inspection area on the display device when the highest matching rate among the plurality of defect candidate data elements related to the inspection area is equal to or greater than a predetermined matching rate threshold, or when the highest other matching rate is equal to or greater than a predetermined matching rate threshold.

[0011] A fourth aspect of the present invention is a display control information generating device of the first aspect (which may be any one of the first to third aspects), wherein the display control information includes control information for causing the display device to display the plurality of defect candidates in the inspection area in order of highest matching rate.

[0012] A fifth aspect of the present invention is the display control information generating device of the first aspect (which may be any one of the first to fourth aspects), wherein each of the plurality of defect candidate data elements further includes another matching rate indicating the degree to which the one defect candidate matches another defect type different from the defect type. The display control information includes control information for causing the display device to display the plurality of defect candidates in the inspection area in descending order of the matching rate and / or the other matching rate.

[0013] A sixth aspect of the present invention is a visual inspection device comprising a display control information generating device according to any one of aspects 1 to 5, a display device, and a display control unit that controls the display device based on the display control information generated by the display control information generating device to cause the display device to display defect candidates on the substrate.

[0014] A seventh aspect of the present invention is a substrate inspection system, which includes a primary inspection device that performs a primary inspection on a substrate, a secondary inspection device that performs a secondary inspection, and the visual inspection device of the sixth aspect.

[0015] Aspect 8 of the present invention is a substrate inspection system comprising a primary inspection device that performs a primary inspection on a substrate, a secondary inspection device that performs a secondary inspection, a display control information generating device of any one of aspects 1 to 5, a display device, and a visual inspection device including a display control unit that controls the display device based on the display control information generated by the display control information generating device to cause the display device to display defect candidates on the substrate.

[0016] A ninth aspect of the present invention is a display control information generation method for generating display control information for displaying defect candidates on a substrate on a display device, the method comprising: (a) storing defect candidate information including a plurality of defect candidate data elements corresponding to a plurality of defect candidates in an inspection area on the substrate; and (b) generating display control information for the plurality of defect candidates in the inspection area based on the defect candidate information. Each of the plurality of defect candidate data elements includes a defect candidate position indicating the position on the substrate of a defect candidate detected by a second inspection using a trained model created by machine learning from tentative defect candidates detected by a rule-based first inspection of the substrate, and a matching rate indicating the degree to which the defect candidate matches a predetermined defect type. The display control information includes control information for hiding the plurality of defect candidates in the inspection area on the display device if the highest matching rate among the plurality of defect candidate data elements related to the inspection area is equal to or greater than a predetermined matching rate threshold.

[0017] A tenth aspect of the present invention provides a computer-readable program for causing a computer to generate display control information for displaying defect candidates on a substrate on a display device, the program causing the computer to execute the following steps: (a) storing defect candidate information including a plurality of defect candidate data elements corresponding to a plurality of defect candidates in an inspection area on the substrate; and (b) generating display control information for the plurality of defect candidates in the inspection area based on the defect candidate information. Each of the plurality of defect candidate data elements includes a defect candidate position indicating the position on the substrate of a defect candidate detected by a second inspection using a trained model created by machine learning from tentative defect candidates detected by a rule-based first inspection of the substrate, and a matching rate indicating the degree to which the defect candidate matches a predetermined defect type. The display control information includes control information for hiding the plurality of defect candidates in the inspection area on the display device when the highest matching rate among the plurality of defect candidate data elements related to the inspection area is equal to or greater than a predetermined matching rate threshold. [Effects of the Invention]

[0018] In the present invention, the time required for the verify operation can be shortened. [Brief explanation of the drawings]

[0019] [Figure 1] 1 is a diagram showing a configuration of a substrate inspection system according to an embodiment; [Figure 2] FIG. [Figure 3] FIG. 2 is a diagram schematically illustrating a configuration of defect candidate information. [Figure 4] FIG. 10 is a diagram illustrating the configuration of a third computer. [Figure 5] FIG. 10 is a block diagram showing functions implemented by a third computer. [Figure 6] FIG. 10 is a diagram showing a flow of generating display control information. [Figure 7] FIG. 10 is a diagram showing a flow of generating display control information. [Figure 8] FIG. 2 is a diagram schematically illustrating a configuration of defect candidate information. [Figure 9] FIG. 2 is a diagram schematically illustrating a configuration of display control information. [Figure 10] FIG. 2 is a diagram schematically illustrating a configuration of display control information. [Figure 11] FIG. 10 is a diagram showing the configuration of another substrate inspection system. DETAILED DESCRIPTION OF THE INVENTION

[0020] 1 is a diagram showing the configuration of a substrate inspection system 1 according to one embodiment of the present invention. The substrate inspection system 1 is a device that inspects the appearance of a substantially flat substrate 9 (also called a panel, board, workpiece, etc.) on which a plurality of pieces intended to become printed wiring boards are mounted. A pattern (for example, a copper wiring pattern or electrode pattern) is formed on the surface of the substrate 9.

[0021] FIG. 2 is a plan view showing an example of a substrate 9. A plurality of sheets 93 are attached to the main surface of the substrate 9, and a plurality of pieces 94 are attached to each sheet 93. The substrate 9, each sheet 93, and each piece 94 are generally rectangular in shape in a plan view. The shapes of the sheets 93 are generally the same, and the shapes of the pieces 94 are also generally the same. In the example shown in FIG. 2, the number of sheets 93 attached to the substrate 9 is four, and the number of pieces 94 attached to each sheet 93 is six. The four sheets 93 are arranged in a matrix, and the six pieces 94 on each sheet 93 are also arranged in a matrix.

[0022] In the inspection of the substrate 9 described below, one piece 94 is set as an "inspection area," which is a unit area for inspection, and a collection of multiple inspection areas included in one sheet 93 is set as an "inspection area group." The shapes of the substrate 9, each sheet 93, and each piece 94 may be changed in various ways. Furthermore, the number of sheets 93 imposed on the substrate 9 and the number of pieces 94 imposed on each sheet 93 may also be changed in various ways within one or more ranges.

[0023] As shown in FIG. 1, the substrate inspection system 1 includes a primary inspection device 2, a secondary inspection device 3, and a visual inspection device 4. The primary inspection device 2 is a device that performs a rule-based primary inspection on a substrate 9 to detect "tentative defect candidates." The primary inspection device 2 sends tentative defect candidate images that include the detected tentative defect candidates to the secondary inspection device 3. The secondary inspection device 3 is a device that performs a secondary inspection (i.e., an AI inspection) on the tentative defect candidate images received from the primary inspection device 2 using a trained model created by machine learning to detect "defect candidates."

[0024] The visual inspection device 4 is a device (so-called verify device) that images the defect candidates detected by the secondary inspection device 3 on the substrate 9 and displays them on a display device such as a display. An operator visually observes the defect candidates displayed on the display device of the visual inspection device 4 and determines whether the defect candidates are fatal true defects that are likely to cause quality problems, or false reports that are virtually unlikely to cause quality problems (i.e., non-fatal normal defects).

[0025] The primary inspection device 2 includes an image acquisition device 21 that captures an image of the substrate 9, and a first computer 22. The image acquisition device 21 includes a stage 211, a stage movement mechanism 212, and an imaging unit 213. The stage 211 is a holder that holds the substrate 9. The stage 211 includes, for example, a vacuum chuck that sucks the lower surface of the substrate 9 and holds it in a horizontal state. The stage movement mechanism 212 is composed of a ball screw, guide rails, a motor, etc., and moves the stage 211 relatively to the imaging unit 213.

[0026] The imaging unit 213 emits illumination light toward the substrate 9 and receives reflected light from the substrate 9 using an imaging sensor or the like, thereby capturing an image of the substrate 9. This image is, for example, a grayscale image of the entire area to be inspected on the substrate 9, and will also be referred to as a "substrate image" in the following description. The substrate image captured by the imaging unit 213 may be, for example, a multi-tone color image. Alternatively, the substrate image may be an image of a portion of the area to be inspected on the substrate 9.

[0027] The first computer 22 has the configuration of a typical computer system including a CPU, GPU, ROM, RAM, fixed disk, display, keyboard, mouse, reading device, communication unit, etc. In the primary inspection device 2, the first computer 22 controls the stage moving mechanism 212 and the imaging unit 213, thereby moving the stage 211 in the horizontal direction and capturing an image of a desired area on the substrate 9. The substrate image captured by the imaging unit 213 is sent to the first computer 22 and stored in the first computer 22.

[0028] The first computer 22 performs a primary inspection on the substrate image, and detects tentative defect candidates. In the primary inspection, for example, an image showing a portion of the substrate 9 (i.e., a portion of the substrate image) is the object of one inspection. Then, the inspection is repeated over the entire area of ​​the substrate 9 to be inspected, thereby performing a primary inspection of the entire substrate image. In the following description, the image that is the object of one inspection in the primary inspection (i.e., an image showing a portion of the substrate 9) is also referred to as the "image to be inspected."

[0029] As described above, the primary inspection is performed by rule-based inspection. Rule-based inspection is an inspection method in which certain rules are set for the shape and other properties of a pattern on the substrate 9, and portions of the pattern on the substrate 9 that do not conform to the rules are determined to be defects (tentative defect candidates in this embodiment), and various methods for this rule-based inspection are known. In the primary inspection, for example, a predetermined feature amount is calculated from each portion of the pattern included in the image to be inspected by a judgment algorithm created in advance based on design rules, and the feature amount is compared with a threshold value for distinguishing between a non-defective product and a tentative defect candidate. Then, based on the comparison result between the feature amount and the threshold value, it is determined whether the image to be inspected is a tentative defect candidate image that includes tentative defect candidates (i.e., whether or not there are tentative defect candidates in the image to be inspected). The feature amount is calculated based on, for example, the line width of linear pattern elements included in the pattern, the end shape of the pattern elements, and the like. Note that in the primary inspection, predetermined image processing may be performed on the image to be inspected before the feature amount is extracted.

[0030] An inspection image that is determined to include a tentative defect candidate in the primary inspection is sent from the primary inspection device 2 to the secondary inspection device 3 and stored in the secondary inspection device 3 as a tentative defect candidate image. In the primary inspection device 2, for example, each time a tentative defect candidate image is detected, the tentative defect candidate image is transmitted to the secondary inspection device 3. In the secondary inspection device 3, multiple tentative defect candidate images are stored as data (hereinafter also referred to as "tentative defect candidate group data") arranged in the order in which they were detected in the primary inspection device 2. In other words, the tentative defect candidate group data includes multiple tentative defect candidate images arranged in the order in which they were detected in the primary inspection device 2.

[0031] Alternatively, in the primary inspection device 2, a plurality of provisional defect candidate images detected in the primary inspection may be stored in the first computer 22 as provisional defect candidate group data arranged in the order in which they were detected. In this case, the provisional defect candidate group data is transmitted from the primary inspection device 2 to the secondary inspection device 3 after the primary inspection of all the images to be inspected is completed.

[0032] The secondary inspection device 3 includes a second computer 32. The second computer 32 has the configuration of a typical computer system including a CPU, a GPU, a ROM, a RAM, a fixed disk, a display, a keyboard, a mouse, a reading device, a communication unit, etc. The provisional defect candidate group data (i.e., a plurality of provisional defect candidate images) sent from the primary inspection device 2 to the secondary inspection device 3 is stored in the second computer 32.

[0033] The second computer 32 performs a secondary inspection on the multiple provisional defect candidate images included in the provisional defect candidate group data, and defect candidates are detected. In the secondary inspection, as described above, an AI inspection is performed using a trained model created by machine learning. The AI ​​inspection is an inspection method that analyzes an image of a pattern on the substrate 9 using an AI (artificial intelligence) image recognition technology and determines the presence or absence of defects (in this embodiment, defect candidates). In addition to the AI ​​inspection, the secondary inspection may also include inspections using other methods.

[0034] The trained model used in the secondary inspection is created, for example, by performing machine learning on the initial model to learn the relationship between an image showing a portion of the substrate 9 (i.e., an image corresponding to the above-mentioned inspection image) and the presence or absence of one defect type. The machine learning is performed, for example, by deep learning using a neural network. The learning by deep learning is performed, for example, by using ResNet. Note that the machine learning may also be performed by a method other than deep learning. The trained model is, for example, generated and stored in advance by the second computer 32.

[0035] In the secondary inspection by the second computer 32, for multiple tentative defect candidate images included in the tentative defect candidate group data, the possibility that the tentative defect candidate included in each tentative defect candidate image is a fatal true defect is examined in the above-mentioned detection order. Specifically, the tentative defect candidate image is input to a trained model, and the matching rate of the trained model for the tentative defect candidate included in the tentative defect candidate image is output. The matching rate means the probability that the tentative defect candidate is the defect type corresponding to the trained model (i.e., the defect type that is a true defect). In other words, the matching rate indicates the degree to which the tentative defect candidate matches the defect type. The matching rate is, for example, a value greater than or equal to 0% and less than or equal to 100%. The higher the matching rate, the higher the possibility that the tentative defect candidate is a true defect, and the lower the matching rate, the lower the possibility that the tentative defect candidate is a true defect.

[0036] In the secondary inspection, if the matching rate of the tentative defect candidate output by the trained model is equal to or greater than a predetermined value (i.e., if the possibility that the candidate is a real defect is high enough), the tentative defect candidate is classified as a defect candidate. Then, a defect candidate data element is generated, which includes the matching rate of the defect candidate and a defect candidate position (e.g., coordinates in an XY coordinate system set on the substrate 9) indicating the position of the defect candidate on the substrate 9. The defect candidate data element is stored in the storage unit of the second computer 32 as one data element constituting the defect candidate information. On the other hand, if the matching rate of the tentative defect candidate output by the trained model is less than the predetermined value, the tentative defect candidate is classified as a false report (i.e., a non-fatal normal defect). Data related to a tentative defect candidate classified as a false report is not included in the defect candidate information.

[0037] The defect candidate information includes a plurality of defect candidate data elements corresponding to the plurality of defect candidates detected in the secondary inspection. In addition to the matching rate and defect candidate position of the corresponding defect candidate, each defect candidate data element further includes detected piece information (i.e., inspection area information) indicating in which piece 94 on the substrate 9 the defect candidate was detected, and detected sheet information (i.e., inspection area group information) indicating in which sheet 93 the piece 94 is included. When the secondary inspection is completed for the plurality of tentative defect candidate images included in the tentative defect candidate group data, the above-mentioned defect candidate information is sent from the secondary inspection device 3 to the visual inspection device 4.

[0038] FIG. 3 is a diagram showing a schematic configuration of defect candidate information (the same applies to FIG. 8 described later). Each of the multiple circles in FIG. 3 indicates a defect candidate data element DE included in the defect candidate information. Each of the multiple rectangles labeled P1 to P24 in FIG. 3 corresponds to the 24 pieces 94 (see FIG. 2) mounted on the substrate 9. In the following description, rectangle P1 will also be referred to as "piece P1," and rectangle P2 will also be referred to as "piece P2." The same applies to rectangles P3 to P24.

[0039] The multiple defect candidate data elements DE included in piece P1 correspond to multiple defect candidates detected in one piece 94 corresponding to piece P1 on the substrate 9 shown in FIG. 2. The numbers inside the circles indicating the defect candidate data elements DE indicate the order in which the defect candidates were detected by the secondary inspection device 3 for the piece 94 corresponding to piece P1 (i.e., the detection order). In the example shown in FIG. 3, piece P1 includes nine defect candidate data elements DE. Piece P2 includes four defect candidate data elements DE. The same is true for the defect candidate data elements DE included in each of pieces P3 to P24. Note that the number of defect candidate data elements DE included in each piece is not limited to the example shown in FIG. 3.

[0040] In FIG. 3, the multiple rectangles labeled ST1 to ST4 correspond to the four sheets 93 (see FIG. 2) imposed on the substrate 9. In the following description, rectangle ST1 will also be referred to as "sheet ST1," and rectangle ST2 will also be referred to as "sheet ST2." The same applies to rectangles ST3 to ST4. Pieces P1 to P6 included in sheet ST1 correspond to six pieces 94 included in one sheet 93 corresponding to sheet ST1 on the substrate 9 shown in FIG. 2. The same applies to pieces P7 to P12 included in sheet ST2, pieces P13 to P18 included in sheet ST3, and pieces P19 to P24 included in sheet ST4.

[0041] In the inspection by the visual inspection device 4 described later, in the piece P1, visual inspection of defect candidates corresponding to the defect candidate data elements DE is performed sequentially from left to right in Fig. 3. Then, when inspection of all the defect candidate data elements DE included in the piece P1 is completed, visual inspection of defect candidates corresponding to the defect candidate data elements DE is performed sequentially from left to right in Fig. 3 in the piece P2 below. Thereafter, visual inspection of the corresponding defect candidates is similarly performed for the defect candidate data elements DE included in the pieces P3 to P24.

[0042] The visual inspection apparatus 4 shown in FIG. 1 includes an imaging device 41 that images a substrate 9, and a third computer 42. The imaging device 41 includes a stage 411, a stage moving mechanism 412, and an imaging unit 413. The stage 411 is a holding unit that holds the substrate 9. The substrate 9 that has undergone primary inspection in the primary inspection apparatus 2 is carried into the visual inspection apparatus 4 and held by the stage 411. The stage 411 includes, for example, a vacuum chuck that sucks the lower surface of the substrate 9 and holds it in a horizontal state. The stage moving mechanism 412 is composed of a ball screw, guide rails, a motor, etc., and moves the stage 411 relatively to the imaging unit 413.

[0043] The imaging unit 413 emits illumination light toward the substrate 9 and receives reflected light from the substrate 9 using an imaging sensor or the like, thereby capturing an image of the substrate 9. This image corresponds to a portion of the substrate 9, and will also be referred to as an "observation image" in the following description. The observation image is, for example, a multi-tone color image. Note that the observation image may also be a grayscale image.

[0044] 4 is a diagram showing the configuration of the third computer 42. The third computer 42 has the configuration of a typical computer system including a CPU 81, a GPU 82, a ROM 83, a RAM 84, a fixed disk 85, a display 86, an input unit 87, a reading device 88, a communication unit 89, and a bus 80. The CPU 81 performs various types of arithmetic processing. The GPU 82 performs various types of arithmetic processing related to image processing. The ROM 83 stores basic programs. The RAM 84 stores various types of information. The fixed disk 85 stores information. The display 86 is a display device that displays various types of information such as images.

[0045] The input unit 87 includes a keyboard 87a and a mouse 87b that accept input from an operator. The reading device 88 reads information from a computer-readable recording medium 881, such as an optical disk, a magnetic disk, a magneto-optical disk, or a memory card. The display 86, the keyboard 87a, the mouse 87b, and the reading device 88 are connected to the bus 80 via an interface I / F. The communication unit 89 transmits and receives signals to and from devices external to the third computer 42. The bus 80 is a signal circuit that connects the CPU 81, the GPU 82, the ROM 83, the RAM 84, the fixed disk 85, the display 86, the input unit 87, the reading device 88, and the communication unit 89.

[0046] In the third computer 42, a program 882 is read in advance from a recording medium 881 via a reading device 88 and stored on a fixed disk 85. The computer-readable program 882 may be stored on the fixed disk 85 via a network. The CPU 81 and the GPU 82 execute arithmetic processing in accordance with the program 882 while using the RAM 84 and the fixed disk 85. The CPU 81 and the GPU 82 function as a calculation unit in the third computer 42. Other components that function as a calculation unit may be employed in addition to the CPU 81 and the GPU 82.

[0047] Fig. 5 is a block diagram showing functions realized by the third computer 42 shown in Fig. 4. Fig. 5 also shows configurations other than the third computer 42. The third computer 42 includes a storage unit 401, a control information generation unit 402, and a display control unit 403. The storage unit 401 is realized by the RAM 84, the fixed disk 85, etc. The control information generation unit 402 and the display control unit 403 are realized by the CPU 81, the GPU 82, the ROM 83, the RAM 84, the fixed disk 85, and their peripheral configurations.

[0048] The storage unit 401 stores the above-mentioned defect candidate information sent from the secondary inspection device 3. The control information generation unit 402 generates display control information for displaying defect candidates on the substrate 9 on the display 86 based on the defect candidate information. The display control information includes control information regarding whether or not to display multiple defect candidates (hereinafter also referred to as "multiple defect candidates included in the defect candidate information") corresponding to multiple defect candidate data elements of the defect candidate information. In other words, the display control information includes control information for hiding defect candidates that do not need to be displayed (i.e., that do not need to be subject to visual inspection by an operator, as described below).

[0049] The display control unit 403 controls the imaging device 41, the display 86, etc. based on the display control information generated by the control information generation unit 402, and causes the display 86 to display defect candidates on the substrate 9. Specifically, for one defect candidate that needs to be displayed in the display control information, the stage moving mechanism 412 is driven based on the defect candidate position of the defect candidate data element corresponding to that defect candidate, and the substrate 9 is positioned so that the defect candidate position is included in the imaging range of the imaging unit 413. Then, the imaging unit 413 images a portion of the substrate 9 that includes the defect candidate, and the acquired observation image is displayed on the display 86.

[0050] An operator of the visual inspection device 4 visually observes the observation image containing the defect candidate and determines whether the defect candidate is a true defect. If the defect candidate contained in the observation image is a true defect, the piece 94 containing the defect candidate (i.e., the true defect) is discarded. Specifically, for example, a mark indicating the discard is applied to the piece 94 in which a true defect is detected. For example, the discarded piece 94 is not subjected to the mounting process of electronic components or the like in a later process.

[0051] Here, assuming a comparative example visual inspection device in which the control information generation unit 402 is omitted, in the comparative example visual inspection device, all of the multiple defect candidates included in the defect candidate information are displayed on the display in sequence in the order in which the defect candidates were detected by the secondary inspection device 3. In substantially the same manner as above, the operator visually observes the observation image and determines whether the defect candidate is a true defect, and if the defect candidate is a true defect, the piece 94 including the defect candidate (i.e., the true defect) is discarded.

[0052] In the visual inspection device of the comparative example, before the true defect is detected, other defect candidates of the piece 94 to be discarded may have been displayed on the display, and the operator may have visually inspected the other defect candidates displayed on the display. For example, if the operator determines that the defect candidate corresponding to the fifth defect candidate data element DE from the left in the piece P1 of the defect candidate information shown in FIG. 3 is a true defect, even though the piece P1 is to be discarded, the four defect candidates corresponding to the first to fourth defect candidate data elements DE from the left of the piece P1 have already been displayed on the display and visually inspected by the operator. Considering that the piece 94 is to be discarded, the work performed on these other defect candidates is unnecessary.

[0053] In contrast, in the visual inspection device 4 according to this embodiment, the control information generator 402 generates display control information related to the display of the defect candidates before the defect candidates are displayed on the display 86. FIGS. 6 and 7 are diagrams showing the flow of generating the display control information. When the display control information is generated, first, the defect candidate information sent from the secondary inspection device 3 to the visual inspection device 4 is stored in the storage unit 401 (see FIG. 5) of the third computer 42 (step S11). Next, based on the defect candidate information (see FIG. 3), the control information generator 402 generates display control information related to the defect candidates included in the plurality of pieces P1 to P24 (step S12).

[0054] In step S12, specifically, the control information generating unit 402 selects one piece from the pieces P1 to P24 of the defect candidate information shown in FIG. 3, and compares the above-mentioned precision rates of the multiple defect candidate data elements DE included in the selected piece. Then, the highest precision rate among the multiple defect candidate data elements DE related to the selected piece (hereinafter also referred to as the "highest precision rate within the piece") is compared with a predetermined precision rate threshold pre-stored in the storage unit 401. The precision rate threshold is, for example, 99%. The control information generating unit 402 compares the highest precision rate within the piece with the precision rate threshold for each of the pieces P1 to P24 of the defect candidate information. Note that the precision rate threshold may be changed as appropriate.

[0055] 8 is a diagram in which the defect candidate data element DE having the highest matching rate in a piece and whose highest matching rate in the piece is equal to or greater than the matching rate threshold is painted black in the defect candidate information shown in FIG. 8. In the example shown in FIG. 8, in sheet ST1, the matching rate of the defect candidate data element DE fifth from the left in piece P1 is equal to or greater than the matching rate threshold. In sheet ST2, the matching rates of the defect candidate data element DE sixth from the left in piece P8, the defect candidate data element DE fourth from the left in piece P10, and the defect candidate data element DE first from the left in piece P11 are each equal to or greater than the matching rate threshold.

[0056] In sheet ST3, the precision rates of the fifth defect candidate data element DE from the left of piece P16 and the fifth defect candidate data element DE from the left of piece P17 are each equal to or greater than the precision rate threshold. In sheet ST4, the precision rates of the second defect candidate data element DE from the left of piece P19, the fourth defect candidate data element DE from the left of piece P21, the sixth defect candidate data element DE from the left of piece P23, and the seventh defect candidate data element DE from the left of piece P24 are each equal to or greater than the precision rate threshold. In the other pieces P2 to P7, P9, P12 to P15, P18, P20, and P22, the highest precision rate within a piece is less than the precision rate threshold.

[0057] The control information generating unit 402 generates control information for a piece whose highest matching rate in the piece is equal to or higher than the matching rate threshold, to hide defect candidates corresponding to all defect candidate data elements DE included in the piece on the display 86, and includes the generated control information in the above-mentioned display control information. In other words, the control information generating unit 402 hides pieces including defect candidates with high matching rates (step S21).

[0058] FIG. 9 is a diagram schematically illustrating the configuration of the display control information (the same applies to FIG. 10, which will be described later). In FIG. 9, the defect candidate data elements DE corresponding to the defect candidates hidden by the control information generating unit 402 are indicated by dashed lines. In the example illustrated in FIG. 9, the defect candidates corresponding to all the defect candidate data elements DE included in the piece P1 on the sheet ST1, the pieces P8, P10, and P11 on the sheet ST2, the pieces P16 and P17 on the sheet ST3, and the pieces P19, P21, P23, and P24 on the sheet ST4 are hidden. Note that the defect candidates corresponding to the defect candidate data elements DE included in the other pieces P2 to P7, P9, P12 to P15, P18, P20, and P22 are not hidden.

[0059] As described above, when the matching rate of a defect candidate data element DE is high, the defect candidate corresponding to the defect candidate data element DE is likely to be a true defect, and the piece that is the inspection area including the defect candidate data element DE is likely to be discarded. Therefore, a piece whose highest matching rate within a piece is equal to or greater than the matching rate threshold is very likely to be discarded, and it is very likely that displaying on the display 86 or visually inspecting the defect candidate corresponding to the defect candidate data element DE included in the piece (hereinafter simply referred to as the "defect candidate included in the piece") would be pointless. Therefore, as described above, the visual inspection device 4 does not display the defect candidate of a piece whose highest matching rate within a piece is equal to or greater than the matching rate threshold, thereby omitting work that is very likely to be pointless and shortening the time required for visual inspection.

[0060] In the visual inspection device 4, if defect candidates are hidden in a predetermined number of pieces (e.g., more than half of the pieces) among the pieces included in one sheet in step S21, the entire sheet may be discarded. In this case, the control information generation unit 402 generates control information for hiding all defect candidates in all pieces included in the sheet (i.e., all inspection areas included in one inspection area group) on the display 86 when defect candidates are hidden in a predetermined number or more (e.g., four or more) of the pieces included in the sheet, and includes the generated control information in the display control information. Similarly, control information is generated for other sheets and included in the display control information. In other words, the control information generation unit 402 hides sheets with a large number of hidden pieces (step S22).

[0061] For example, in the example shown in FIG. 9, the number of pieces hidden in sheets ST1 to ST3 is less than four, and the number of pieces hidden in sheet ST4 is four (i.e., four or more). Therefore, as shown in FIG. 10, defect candidates corresponding to all defect candidate data elements DE are hidden in all pieces P19 to P24 included in sheet ST4. This eliminates the need to display on the display 86 or perform visual inspection for defect candidates corresponding to defect candidate data elements DE included in sheet ST4 to be discarded, thereby shortening the time required for visual inspection. Note that in FIG. 10, the display / hide status of pieces P1 to P18 on sheets ST1 to ST3 is the same as that shown in FIG. 9.

[0062] The control information generator 402 may also generate control information for changing the display order of multiple defect candidates included in each piece P2 to P7, P9, P12 to P15, and P18 for which the defect candidates were not hidden in steps S21 and S22. Specifically, the control information generator 402 changes the order of multiple defect candidate data elements DE (i.e., the order of the defect candidates) in each of the pieces P2 to P7, P9, P12 to P15, and P18 so that the order is from highest to lowest matching rate (step S23).

[0063] As a result, when the operator visually inspects the defect candidates contained in each piece, defect candidates that are relatively likely to be determined to be true defects are displayed on the display 86 before defect candidates that are relatively unlikely to be determined to be true defects. This reduces the possibility that other defect candidates contained in the piece will be displayed on the display 86 and that the operator will visually inspect the other defect candidates displayed on the display 86 before the operator detects a true defect and discards the piece containing the true defect. As a result, it is possible to omit work that is likely to be wasted and shorten the time required for visual inspection.

[0064] In the above description, the defect candidate data element of the defect candidate information generated by the secondary inspection apparatus 3 includes a match rate of the defect candidate for one defect type. However, it may also include match rates of the defect candidates for multiple defect types. In other words, the defect candidate data element may include multiple match rates. In this case, for example, multiple trained models corresponding to multiple defect types are prepared in the secondary inspection apparatus 3. Then, the provisional defect candidate images included in the provisional defect candidate group data generated by the primary inspection apparatus 2 are input to the multiple trained models, respectively, and the match rates of the multiple trained models are output for the provisional defect candidates included in the provisional defect candidate images. Thereafter, when the provisional defect candidates are classified into defect candidates based on the multiple match rates of the provisional defect candidates, the multiple match rates are included in the defect candidate data element related to the defect candidate.

[0065] In this way, when one defect candidate data element includes multiple precision rates corresponding to multiple defect types, in step S21, the control information generation unit 402 finds multiple highest precision rates within the piece corresponding to the multiple defect types, respectively. Then, each of the multiple highest precision rates within the piece is compared with a precision threshold value for the corresponding defect type. The precision threshold value may be different for each defect type, for example, or may be the same regardless of the defect type. When at least one highest precision rate within the piece among the multiple highest precision rates within the piece is equal to or greater than the corresponding precision threshold value, the control information generation unit 402 generates control information for hiding all defect candidates included in the piece, and includes the control information in the display control information.

[0066] Furthermore, when one defect candidate data element includes multiple matching rates corresponding to multiple defect types, in step S23, the defect candidates in each piece are sorted in descending order of one or more of the multiple matching rates. Specifically, for example, the defect candidates in each piece are sorted in descending order of one matching rate selected from the multiple matching rates. Alternatively, the defect candidates in each piece may be sorted in descending order of the maximum value of two or more matching rates associated with each defect candidate. Furthermore, the defect candidates in each piece may be sorted in descending order of the average value of two or more matching rates associated with each defect candidate.

[0067] In the visual inspection device 4, as described above, the display control unit 403 controls the imaging device 41, the display 86, etc. based on the display control information generated by the control information generation unit 402, thereby sequentially displaying only the defect candidates that have not been hidden (i.e., those that have been determined to need to be displayed) on the display 86. This reduces the number of opportunities for displaying on the display 86 and for the operator to visually inspect the defect candidates contained in the pieces 94 and sheets 93 to be discarded on the substrate 9 shown in FIG. 2. In other words, it reduces the number of verification operations for the defect candidates contained in the pieces 94 and sheets 93 to be discarded. As a result, it is possible to shorten the time required for the verification operation.

[0068] In the visual inspection device 4 described above, a display control information generating device 400 that generates display control information is configured by a storage unit 401 and a control information generating unit 402 shown in Fig. 5. When the display control information generating device 400 focuses on one piece 94, the defect candidate information stored in the storage unit 401 includes a plurality of defect candidate data elements that respectively correspond to a plurality of defect candidates in the piece 94, which is an inspection area. Furthermore, the control information generating unit 402 generates display control information related to the plurality of defect candidates in the piece 94, which is an inspection area, based on the defect candidate information.

[0069] As described above, the display control information generating device 400 is a device that generates display control information for displaying defect candidates on the substrate 9 on a display device (i.e., the display 86). The display control information generating device 400 includes a storage unit 401 and a control information generating unit 402. The storage unit 401 stores defect candidate information. The defect candidate information includes a plurality of defect candidate data elements that respectively correspond to a plurality of defect candidates in an inspection area on the substrate 9 (piece 94 in the above example). The control information generating unit 402 generates display control information related to the plurality of defect candidates in the inspection area based on the defect candidate information.

[0070] Each of the plurality of defect candidate data elements includes a defect candidate location and a matching rate. The defect candidate location indicates the location on the substrate 9 of a defect candidate detected by a secondary inspection using a trained model created by machine learning from tentative defect candidates detected by a rule-based primary inspection of the substrate 9. The matching rate indicates the degree to which the defect candidate matches a predetermined defect type. The display control information includes control information for hiding the plurality of defect candidates in the inspection area on the display 86 when the highest matching rate (i.e., the highest matching rate within the piece) among the plurality of defect candidate data elements related to the inspection area is equal to or greater than a predetermined matching rate threshold. This allows inspection areas that are likely to be discarded to be excluded from the verification process, as described above, thereby shortening the time required for the verification process.

[0071] As described above, an inspection area group, which is a plurality of inspection areas (pieces 94 in the above example) including the above inspection area, is set on the substrate 9. The defect candidate information includes a plurality of defect candidate data elements corresponding to a plurality of defect candidates in each inspection area of ​​the inspection area group. Each of the plurality of defect candidate data elements includes a defect candidate position and a matching rate. The display control information includes control information for hiding the plurality of defect candidates in each inspection area on the display 86 when the highest matching rate (the highest matching rate in the piece in the above example) among the plurality of defect candidate data elements for each inspection area is equal to or greater than a matching rate threshold. Preferably, the display control information includes control information for hiding all defect candidates included in the inspection area group on the display 86 when a predetermined number or more of the plurality of inspection areas (i.e., the plurality of inspection areas included in the inspection area group) have had their defect candidates hidden. This makes it possible to exclude inspection area groups that are likely to be discarded from the verification process, as described above, and further shorten the time required for the verification process.

[0072] As described above, each of the plurality of defect candidate data elements further includes an additional matching rate. The additional matching rate indicates the degree to which a defect candidate matches another defect type different from the defect type. Preferably, the display control information includes control information for hiding the plurality of defect candidates in the inspection area on the display 86 when the highest matching rate among the plurality of defect candidate data elements associated with the inspection area is equal to or greater than a predetermined matching rate threshold, or when the highest other matching rate is equal to or greater than a predetermined matching rate threshold. This allows inspection areas that are likely to be discarded to be excluded from the verification process, taking into account the matching rates of the defect candidates for the plurality of defect types, as described above. As a result, the time required for the verification process can be shortened.

[0073] As described above, the display control information preferably includes control information for displaying multiple defect candidates in the inspection area on the display 86 in descending order of relevance. This allows true defects, if any, to be detected at a relatively early stage in the verification process for the inspection area. This further reduces the time required for the verification process.

[0074] As described above, each of the plurality of defect candidate data elements further includes another matching rate. The other matching rate indicates the degree to which a defect candidate matches another defect type different from the defect type. Preferably, the display control information includes control information for displaying the plurality of defect candidates in the inspection area on the display device in descending order of the matching rate and / or the other matching rate. This allows true defects corresponding to any of the plurality of defect types in the inspection area to be detected at a relatively early stage in the verification process for the inspection area. This further reduces the time required for the verification process.

[0075] The visual inspection apparatus 4 described above includes the display control information generating device 400, a display device (i.e., display 86), and a display control unit 403. The display control unit 403 controls the display 86 based on the display control information generated by the display control information generating device 400, and causes the display 86 to display defect candidates on the substrate 9. As described above, the visual inspection apparatus 4 can shorten the time required for the verification work.

[0076] The above-mentioned substrate inspection system 1 includes a primary inspection device 2 that performs a primary inspection on the substrate 9, a secondary inspection device 3 that performs a secondary inspection, and the above-mentioned visual inspection device 4. As described above, the substrate inspection system 1 can shorten the time required for the verification work, and therefore can shorten the time required for inspection of the substrate 9 (i.e., the primary inspection, secondary inspection, and visual inspection).

[0077] The above-mentioned display control information generation method includes a step (step S11) of storing defect candidate information including a plurality of defect candidate data elements each corresponding to a plurality of defect candidates in an inspection area on the substrate 9 (in the above example, piece 94), and a step (step S12) of generating display control information relating to the plurality of defect candidates in the inspection area based on the defect candidate information.

[0078] Each of the plurality of defect candidate data elements includes a defect candidate position and a matching rate. The defect candidate position indicates the position on the substrate 9 of a defect candidate detected by a secondary inspection using a trained model created by machine learning from tentative defect candidates detected by a rule-based primary inspection of the substrate 9. The matching rate indicates the degree to which the defect candidate matches a predetermined defect type. The display control information includes control information for hiding the plurality of defect candidates in the inspection area on the display 86 when the highest matching rate (i.e., the highest matching rate within the piece) among the plurality of defect candidate data elements relating to the inspection area is equal to or greater than a predetermined matching rate threshold. This, as described above, can shorten the time required for the verification operation.

[0079] The above-mentioned computer-readable program 882 causes a computer (in the above example, the third computer 42) to execute the following steps: storing defect candidate information (step S11) including a plurality of defect candidate data elements each corresponding to a plurality of defect candidates in an inspection area on the substrate 9 (in the above example, piece 94); and generating display control information relating to the plurality of defect candidates in the inspection area based on the defect candidate information (step S12).

[0080] Each of the plurality of defect candidate data elements includes a defect candidate position and a matching rate. The defect candidate position indicates the position on the substrate 9 of a defect candidate detected by a secondary inspection using a trained model created by machine learning from tentative defect candidates detected by a rule-based primary inspection of the substrate 9. The matching rate indicates the degree to which the defect candidate matches a predetermined defect type. The display control information includes control information for hiding the plurality of defect candidates in the inspection area on the display 86 when the highest matching rate (i.e., the highest matching rate within the piece) among the plurality of defect candidate data elements relating to the inspection area is equal to or greater than a predetermined matching rate threshold. This, as described above, can shorten the time required for the verification operation.

[0081] In the above description, the display control information generating device 400 is provided in the visual inspection device 4, but this is not limiting. For example, the display control information generating device 400 may be provided in a processing tool 5 provided between the secondary inspection device 3 and the visual inspection device 4a, as in the substrate inspection system 1a shown in FIG. 11 . The substrate inspection system 1a includes a primary inspection device 2, a secondary inspection device 3, a processing tool 5, and a visual inspection device 4a. The visual inspection device 4a has substantially the same configuration as the visual inspection device 4 described above, except that the display control information generating device 400 is not provided. The processing tool 5 receives the defect candidate information output from the secondary inspection device 3, performs processing such as data conversion on the defect candidate information so that it is compatible with processing in the visual inspection device 4a, and then sends the processed defect candidate information to the third computer 42 of the visual inspection device 4a.

[0082] The processing tool 5 includes a fourth computer 52. The fourth computer 52 has a typical computer system configuration including a CPU, a GPU, a ROM, a RAM, a fixed disk, a display, a keyboard, a mouse, a reading device, a communication unit, etc., similar to the third computer 42 shown in Fig. 4. The fourth computer 52 realizes a display control information generating device 400 (see Fig. 5) including the above-described storage unit 401 and control information generating unit 402.

[0083] As described above, the substrate inspection system 1a includes the primary inspection device 2, the secondary inspection device 3, the display control information generating device 400, and the visual inspection device 4a. The primary inspection device 2 performs a primary inspection on the substrate 9. The secondary inspection device 3 performs a secondary inspection. The visual inspection device 4a includes a display device (i.e., the display 86 shown in FIG. 4) and a display control unit 403 (see FIG. 5). The display control unit 403 controls the display 86 based on the display control information generated by the display control information generating device 400, and causes the display 86 to display defect candidates on the substrate 9. In the substrate inspection system 1a, similarly to the substrate inspection system 1, the time required for the verification operation can be shortened, and therefore the time required for inspecting the substrate 9 can be shortened.

[0084] The display control information generating device 400, the visual inspection device 4, 4a, the substrate inspection system 1, 1a, the display control information generating method, and the program 882 described above can be modified in various ways.

[0085] For example, in generating the display control information in step S12, it is not necessary to hide all defect candidates in the sheet in step S22 and / or change the order of the defect candidates in step S23.

[0086] The above-mentioned inspection area does not necessarily have to be the piece 94 imposed on the substrate 9, but may be, for example, the sheet 93 or the substrate 9 itself, as long as it is a unit area for inspection of the substrate 9.

[0087] The visual inspection device 4 does not necessarily have to be included in the above-described substrate inspection system 1, and may be used independently of the primary inspection device 2 and the secondary inspection device 3. Furthermore, the display control information generating device 400 does not necessarily have to be provided in the substrate inspection system 1, 1a, and may be used to generate display control information in a device other than the substrate inspection system 1, 1a.

[0088] The substrate 9 does not necessarily have to be a substrate intended to become a printed wiring board, and may be variously modified.

[0089] The configurations in the above-described embodiment and each modification may be combined as appropriate as long as they are not mutually contradictory. [Explanation of symbols]

[0090] 1,1a PCB inspection system 2 Primary inspection equipment 3 Secondary inspection equipment 4,4a Visual inspection equipment 9 Substrate 86 Display 93 seats 94 pieces 400 Display control information generating device 401 Storage section 402 Control information generation unit 403 Display control unit 882 Programs S11~S12, S21~S23 steps

Claims

1. A display control information generating device that generates display control information for displaying defect candidates on a substrate on a display device, a storage unit that stores defect candidate information including a plurality of defect candidate data elements that respectively correspond to a plurality of defect candidates in an inspection area on a substrate; a control information generating unit that generates display control information related to the plurality of defect candidates in the inspection area based on the defect candidate information; Equipped with Each of the plurality of defect candidate data elements comprises: a defect candidate position indicating the position on the substrate of one defect candidate detected by a secondary inspection using a trained model created by machine learning from tentative defect candidates detected by a rule-based primary inspection of the substrate; and a matching rate indicating the degree to which the one defect candidate matches a predetermined defect type; Including, A display control information generating device, wherein the display control information includes control information for hiding the plurality of defect candidates in the inspection area on a display device when the highest matching rate among the plurality of defect candidate data elements relating to the inspection area is equal to or greater than a predetermined matching rate threshold.

2. 2. The display control information generating device according to claim 1, an inspection area group, which is a plurality of inspection areas including the inspection area, is set on the substrate; the defect candidate information includes a plurality of defect candidate data elements each corresponding to a plurality of defect candidates in each inspection area of ​​the group of inspection areas and each including the defect candidate position and the matching rate; the display control information includes control information for hiding the plurality of defect candidates in each of the inspection areas on the display device when the highest matching rate among the plurality of defect candidate data elements related to each of the inspection areas is equal to or greater than the matching rate threshold; The display control information generating device includes control information that, when the plurality of defect candidates are hidden in a predetermined number or more of the plurality of inspection areas, hides all defect candidates included in the group of inspection areas on the display device.

3. 2. The display control information generating device according to claim 1, each of the plurality of defect candidate data elements further includes another matching ratio indicating a degree to which the one defect candidate matches another defect type different from the defect type; A display control information generating device, wherein the display control information includes control information for hiding the plurality of defect candidates in the inspection area on the display device when the highest matching rate among the plurality of defect candidate data elements relating to the inspection area is equal to or greater than a predetermined matching rate threshold, or when the highest other matching rate is equal to or greater than a predetermined matching rate threshold.

4. 2. The display control information generating device according to claim 1, A display control information generating device, wherein the display control information includes control information for displaying the plurality of defect candidates in the inspection area on the display device in descending order of the matching rate.

5. 2. The display control information generating device according to claim 1, Each of the plurality of defect candidate data elements comprises: Further, another matching ratio indicating a degree to which the one defect candidate matches another defect type different from the defect type; A display control information generating device, wherein the display control information includes control information for causing the display device to display the plurality of defect candidates in the inspection area in descending order of the matching rate and / or the other matching rate.

6. A visual inspection device, comprising: A display control information generating device according to any one of claims 1 to 5; A display device; a display control unit that controls the display device based on the display control information generated by the display control information generating device to display defect candidates on the substrate on the display device; A visual inspection device comprising:

7. 1. A substrate inspection system, comprising: a primary inspection device that performs a primary inspection on the board; a secondary inspection device that performs a secondary inspection; The visual inspection device according to claim 6; A substrate inspection system comprising:

8. 1. A substrate inspection system, comprising: a primary inspection device that performs a primary inspection on the board; a secondary inspection device that performs a secondary inspection; A display control information generating device according to any one of claims 1 to 5; a visual inspection device including a display device and a display control unit that controls the display device based on the display control information generated by the display control information generating device to cause the display device to display defect candidates on the substrate; A substrate inspection system comprising:

9. A display control information generation method for generating display control information for displaying defect candidates on a substrate on a display device, comprising: a) storing defect candidate information including a plurality of defect candidate data elements respectively corresponding to a plurality of defect candidates in an inspection area on a substrate; b) generating display control information related to the plurality of defect candidates in the inspection area based on the defect candidate information; Equipped with Each of the plurality of defect candidate data elements comprises: a defect candidate position indicating the position on the substrate of one defect candidate detected by a secondary inspection using a trained model created by machine learning from tentative defect candidates detected by a rule-based primary inspection of the substrate; and a matching rate indicating the degree to which the one defect candidate matches a predetermined defect type; Including, A display control information generating method, wherein the display control information includes control information for hiding the plurality of defect candidates in the inspection area on a display device when the highest matching rate among the plurality of defect candidate data elements relating to the inspection area is equal to or greater than a predetermined matching rate threshold.

10. A computer-readable program that causes a computer to generate display control information for displaying defect candidates on a substrate on a display device, a) storing defect candidate information including a plurality of defect candidate data elements respectively corresponding to a plurality of defect candidates in an inspection area on a substrate; b) generating display control information related to the plurality of defect candidates in the inspection area based on the defect candidate information; on the computer, Each of the plurality of defect candidate data elements comprises: a defect candidate position indicating the position on the substrate of one defect candidate detected by a secondary inspection using a trained model created by machine learning from tentative defect candidates detected by a rule-based primary inspection of the substrate; and a matching rate indicating the degree to which the one defect candidate matches a predetermined defect type; Including, A computer-readable program, wherein the display control information includes control information for hiding the plurality of defect candidates in the inspection area on a display device when the highest matching rate among the plurality of defect candidate data elements relating to the inspection area is equal to or greater than a predetermined matching rate threshold.

Images