JP7680862B2 - Apparatus and method for supporting the creation of teacher data - Google Patents

Description

The present invention relates to a teacher data creation support device and a teacher data creation support method that support the creation of teacher data.

Products, foods, and other items are inspected appropriately at each stage of production and before distribution to determine whether they are good or not. For example, Patent Document 1 describes a defect inspection device for inspecting various defects that occur during the manufacturing process of semiconductor wafers.

In this defect inspection device, multiple SEM (scanning electron microscope) images are generated that show multiple wafers to be inspected. From the multiple SEM images generated, a SEM image that shows a good wafer is designated by the user as a template. The difference between each of the multiple SEM images (inspection images) other than the template and the template is calculated, and defects in the circuit pattern of the wafer shown by the inspection image are detected.

JP 2011-119471 A

In defect inspection devices, even if an object being inspected is not defective, it may be determined that it is defective. In this case, even though there is no defect, the object being inspected may be discarded due to erroneous inspection results, resulting in a decrease in yield. For this reason, it is desirable to perform inspections with higher accuracy.

The object of the present invention is to provide a teacher data creation support device and a teacher data creation support method that can easily create teacher data for conducting highly accurate inspections.

(1) A teacher data creation support device of a first invention is a teacher data creation support device that supports the creation of teacher data to be used for reinspection of an inspection object, and includes an abnormal data acquisition unit that acquires abnormal data indicating an image of an inspection object that has been determined in advance to have a defect, a normal data acquisition unit that acquires normal data indicating an image of a normal inspection object corresponding to the abnormal data acquired by the abnormal data acquisition unit, a differential data generation unit that generates differential data indicating the difference between the abnormal data acquired by the abnormal data acquisition unit and the normal data acquired by the normal data acquisition unit so as to correspond to the abnormal data, a presentation unit that presents an image of a portion of the inspection object based on the differential data generated by the differential data generation unit , a reception unit that accepts a selection of the differential data presented by the presentation unit, and a registration unit that registers the differential data selected by the reception unit as teacher data .

In this teacher data creation support device, an image of a portion of an inspection object is presented based on difference data that indicates the difference between abnormal data, which shows an image of an inspection object previously determined to have a defect, and normal data, which shows an image of a normal inspection object. The portion of the inspection object in the image shown based on the difference data is likely to require reinspection. Therefore, by checking the presented image, the user can create teacher data for reinspecting that portion. This makes it easy to create teacher data for performing inspections with high accuracy.

Furthermore , the difference data selected by the user is registered as teacher data, which makes it easier to create teacher data.

( 2 ) The receiving unit may further receive corrections to the difference data presented by the presenting unit, and the registering unit may register the corrected difference data as training data. In this case, it is possible to more appropriately correct the portion of the inspection object that requires re-inspection in the image based on the difference data. This makes it possible to create training data for performing inspections with higher accuracy.

( 3 ) The normal data acquisition unit may acquire a plurality of normal data corresponding to the abnormal data, and the differential data generation unit may generate a plurality of differential data based on the abnormal data and the plurality of normal data corresponding to the abnormal data. In this case, a large number of differential data are generated from one abnormal data. This can improve the work efficiency of creating teacher data.

( 4 ) The normal data acquisition unit may acquire multiple normal data corresponding to the abnormal data, and the differential data generation unit may generate differential data based on the abnormal data and an average of the multiple normal data corresponding to the abnormal data. With this configuration, even if a noise component unrelated to a defect accidentally gets mixed into any of the multiple normal data, the multiple normal data are averaged, so that the noise component has almost no effect on the pixel value of the normal data after averaging. Therefore, it is possible to create teacher data for performing inspection with higher accuracy.

( 5 ) The normal data acquisition unit may acquire normal data corresponding to a plurality of abnormal data, and the differential data generation unit may generate differential data based on each abnormal data and the normal data corresponding to the abnormal data. In this case, teacher data can be created at high speed using the common normal data.

( 6 ) The normal data acquired by the normal data acquisition unit may include image data showing an image of an inspection object that has not been determined to have a defect in advance. In this case, the normal data showing an image of a normal inspection object can be easily acquired.

( 7 ) The normal data acquired by the normal data acquisition unit may include master data showing a design drawing of the inspection object. In this case, the normal data showing an image of the normal inspection object can be easily acquired.

( 8 ) The normal data acquisition unit may acquire master data that has been corrected based on the processing accuracy of the object to be inspected as the normal data. With this configuration, even if the area to be inspected is minute, normal data that shows an image of a normal object to be inspected can be easily acquired.

( 9 ) A non-inspection target area may be set in the abnormal data acquired by the abnormal data acquisition unit and the normal data acquired by the normal data acquisition unit, and the difference data generation unit may generate the difference data by excluding the set non-inspection target area. In this case, the image shown by the difference data is prevented from including a portion outside the inspection target area. This makes it possible to create teacher data for performing inspections with higher accuracy.

( 10 ) The abnormality data acquisition unit may further acquire defect information indicating a type of defect in the inspection object corresponding to the acquired abnormality data, and the differential data generation unit may add the defect information acquired by the abnormality data acquisition unit to the generated differential data. In this case, the user does not need to perform the task of adding the defect information to the teacher data. This reduces the burden on the user and improves the work efficiency of creating the teacher data. In addition, since there is no error associated with the user's work, more accurate teacher data can be created.

( 11 ) The abnormal data acquisition unit may acquire binarized abnormal data, and the normal data acquisition unit may acquire binarized normal data. In this case, the amount of abnormal data and normal data is reduced, so that teacher data can be created at high speed.

(12) A teacher data creation support method according to a second aspect of the present invention is a teacher data creation support method that supports the creation of teacher data to be used for reinspecting an inspection object, in which a teacher data creation support device acquires abnormal data indicating an image of an inspection object that has been previously determined to have a defect, acquires normal data indicating an image of a normal inspection object that corresponds to the acquired abnormal data, generates differential data indicating the difference between the acquired abnormal data and normal data acquired to correspond to the abnormal data, presents an image of a portion of the inspection object based on the generated differential data, accepts a selection of the presented differential data, and registers the differential data whose selection has been accepted as teacher data.

According to this method for supporting the creation of teacher data, an image of a portion of an object to be inspected is presented based on differential data indicating the difference between abnormal data indicating an image of an object to be inspected that has been determined in advance to have a defect and normal data indicating an image of a normal object to be inspected. The portion of the object to be inspected in the image shown based on the differential data is likely to require reinspection. Therefore, by checking the presented image, a user can create teacher data for reinspecting that portion. This makes it possible to easily create teacher data for performing inspections with high accuracy. Furthermore, the differential data selected by the user is registered as teacher data. This makes it possible to more easily create teacher data.

The present invention makes it possible to create accurate training data while reducing the burden on users.

1 is a diagram showing a configuration of a processing system including a support device according to a first embodiment of the present invention; FIG. 2 is a diagram showing a configuration of the support device of FIG. 1; FIG. 2 is a diagram showing various data used to create teacher data. FIG. 2 is a diagram showing an example of a display screen of a display device when creating teacher data. 3 is a flowchart showing a support process performed by the support device of FIG. 2 . FIG. 11 is a diagram showing various data used to create teacher data in the first modified example. FIG. 13 is a diagram showing various data used to create teacher data in a second modified example. FIG. 13 is a diagram showing various data used to create teacher data in a third modified example. FIG. 11 is a diagram showing various data used to create teacher data in the second embodiment. FIG. 13 is a diagram showing various data used to create teacher data in the third embodiment. FIG. 13 is a diagram showing various data used to create teacher data in the fourth embodiment.

[1] First embodiment (1) Processing system A teacher data creation support device and a teacher data creation support method according to an embodiment of the present invention will be described below with reference to the drawings. In the following description, the teacher data creation support device will be abbreviated to support device. FIG. 1 is a diagram showing the configuration of a processing system including a support device according to a first embodiment of the present invention. As shown in FIG. 1, the processing system 100 includes a processing device 10, an inspection device 20, and a database storage device 30.

The processing device 10 is composed of a CPU (Central Processing Unit) 11, a RAM (Random Access Memory) 12, a ROM (Read Only Memory) 13, a storage device 14, an operation unit 15, a display device 16, and an input/output I/F (interface) 17. The CPU 11, RAM 12, ROM 13, storage device 14, operation unit 15, display device 16, and input/output I/F 17 are connected to a bus 18.

RAM 12 is used as a working area for CPU 11. ROM 13 stores a system program. Storage device 14 includes a storage medium such as a hard disk or semiconductor memory, and stores a teacher data creation support program (hereinafter abbreviated as the support program). The support program may be stored in ROM 13 or another external storage device. CPU 11, RAM 12, and ROM 13 constitute a support device 40 for executing a teacher data creation support process (hereinafter abbreviated as the support process). In the support process, the creation of teacher data is supported.

The operation unit 15 is an input device such as a keyboard, a mouse, or a touch panel. A user can give predetermined instructions to the support device 40 by operating the operation unit 15. The display device 16 is a display device such as a liquid crystal display device, and displays a GUI (Graphical User Interface) or the like for receiving instructions from the user. The input/output I/F 17 is connected to the inspection device 20.

The inspection device 20 is, for example, an AOI (automated optical inspection) device, which sequentially captures images of the objects to be inspected to generate multiple image data respectively showing images of the objects to be inspected, and stores each of the generated image data. Each stored image data is assigned a unique identification number.

The inspection device 20 will be described below using a substrate as an example of an object to be inspected, but the object to be inspected is not limited to a substrate. Note that a substrate refers to a semiconductor substrate, a substrate for an FPD (Flat Panel Display) such as a liquid crystal display device or an organic EL (Electro Luminescence) display device, a substrate for an optical disk, a substrate for a magnetic disk, a substrate for a magneto-optical disk, a substrate for a photomask, a ceramic substrate, or a substrate for a solar cell, etc.

The inspection device 20 inspects the board corresponding to each image data by processing each stored image data based on a predetermined algorithm. The inspection device 20 may inspect the board corresponding to each image data based on deep learning. In the inspection, it is determined whether or not the board has a defect. Furthermore, for boards determined to have a defect, the type of the defect is determined.

In the inspection device 20, even if a board is not defective, it may be determined that it is defective. If a board is discarded due to an erroneous determination even though there is no defect, the yield will decrease. Therefore, for boards determined to be defective, reinspection is performed using supervised learning. The support device 40 supports the creation of teacher data to be used in the reinspection. The database storage device 30 includes a large-capacity storage device such as a server. The created teacher data is registered in the database storage device 30. Details of the support device 40 are described below.

(2) Support device FIG. 2 is a diagram showing the configuration of the support device 40 in FIG. 1. FIG. 3 is a diagram showing various data used in creating teacher data. FIG. 4 is a diagram showing an example of a display screen of the display device 16 in creating teacher data. As shown in FIG. 2, the support device 40 includes, as functional units, an abnormal data acquisition unit 41, a normal data acquisition unit 42, a difference data generation unit 43, a presentation unit 44, a reception unit 45, and a registration unit 46. The functional units of the support device 40 are realized by the CPU 11 in FIG. 1 executing a support program stored in the ROM 13 or the storage device 14, etc. Some or all of the functional units of the support device 40 may be realized by hardware such as electronic circuits.

The abnormality data acquisition unit 41 acquires image data (hereinafter referred to as abnormality data) showing an image of each board that has been determined in advance by the inspection device 20 to be defective from the inspection device 20. The image data may show an image of the entire board, or may show a partial image of the board if it is the same area. A portion of the image of the board based on the abnormality data acquired by the abnormality data acquisition unit 41 is shown in the upper left corner of Figure 3.

The normal data acquisition unit 42 acquires image data (hereinafter referred to as normal data) showing an image of a normal board so as to correspond to each abnormal data acquired by the abnormal data acquisition unit 41. In this example, the normal data is predetermined image data showing an image of a board that has not been determined in advance to have a defect by the inspection device 20, and is acquired from the inspection device 20. A portion of the image of the board based on the normal data acquired by the normal data acquisition unit 42 is shown in the lower left part of Figure 3.

Abnormal data and normal data correspond to each other. For example, when abnormal data is acquired by the abnormal data acquisition unit 41, image data having an identification number one before the identification number of the abnormal data and showing an image of a normal board may be acquired as normal data by the normal data acquisition unit 42. Alternatively, when abnormal data is acquired by the abnormal data acquisition unit 41, image data having an identification number one after the identification number of the abnormal data and showing an image of a normal board may be acquired as normal data by the normal data acquisition unit 42.

The differential data generation unit 43 generates new image data by calculating the difference between each pixel value between the abnormal data acquired by the abnormal data acquisition unit 41 and the normal data acquired by the normal data acquisition unit 42 so as to correspond to the abnormal data. The image data generated by the differential data generation unit 43 is called differential data. The right part of FIG. 3 shows an image of the board based on the differential data generated by the differential data generation unit 43. The differential data shows an image of a part of the board that is likely to require reinspection. Therefore, the differential data can serve as a label indicating the part of the board that requires reinspection.

The presentation unit 44 presents each differential data to the user by displaying on the display device 16 a GUI 50 ( FIG. 4 ) including an image based on each differential data generated by the differential data generation unit 43. As shown in FIG. 4 , the GUI 50 includes an image display area 51, a registration button 52, and a correction button 53. A plurality of images of the measurement object are displayed in the image display area 51. In this example, the image based on the differential data is displayed in the image display area 51 so as to be superimposed on the image based on the abnormality data, but only the image based on the differential data may be displayed in the image display area 51.

The reception unit 45 receives an instruction to register the difference data. Specifically, the reception unit 45 receives a selection of difference data to be registered as teacher data from the operation unit 15 in the GUI 50 displayed by the presentation unit 44. While visually checking the images displayed in the image display area 51, the user can select an arbitrary image using the operation unit 15 and operate the registration button 52 to give an instruction to the reception unit 45 to select the difference data representing that image as teacher data. The registration unit 46 registers the difference data whose selection has been accepted by the reception unit 45 in the database storage device 30 as teacher data.

The reception unit 45 can also accept corrections to difference data. The user can use the operation unit 15 to select any image in the image display area 51 and operate the correction button 53 to instruct the reception unit 45 to correct the difference data corresponding to that image. The user can also use the operation unit 15 to specify, for example, filling in parts of the selected image that require re-examination.

When the reception unit 45 receives the designation, the differential data generation unit 43 assigns a label indicating the portion for which the designation was received to the selected differential data. This causes the differential data to be corrected. After the differential data has been corrected, when the registration button 52 is operated, the registration unit 46 registers the corrected differential data in the database storage device 30 as teacher data.

(3) Support Processing Fig. 5 is a flowchart showing the support processing by the support device 40 of Fig. 2. The support processing of Fig. 5 is performed by the CPU 11 of Fig. 1 executing on the RAM 12 a support program stored in the ROM 13 or the storage device 14, etc. The support processing will be described below with reference to the support device 40 of Fig. 2 and the flowchart of Fig. 5.

First, the abnormal data acquisition unit 41 acquires each abnormal data from the inspection device 20 (step S1). Next, the normal data acquisition unit 42 acquires normal data corresponding to each abnormal data acquired in step S1 from the inspection device 20 (step S2). Steps S1 and S2 may be executed simultaneously.

Next, the differential data generating unit 43 generates each differential data based on the corresponding abnormal data and normal data acquired in steps S1 and S2 (step S3). After that, the presenting unit 44 presents each differential data to the user by displaying an image based on each differential data generated in step S3 on the display device 16 (step S4).

Next, the reception unit 45 determines whether or not a correction to any of the difference data presented in step S4 has been accepted (step S5). If a correction to any of the difference data has not been accepted, the reception unit 45 proceeds to step S7. If a correction to any of the difference data has been accepted, the difference data generation unit 43 corrects the difference data by adding a label to the difference data indicating the portion for which the correction designation has been accepted (step S6), and proceeds to step S7.

In step S7, the reception unit 45 determines whether or not an instruction has been given to register any of the difference data presented in step S4 or the difference data corrected in step S6 (step S7). If an instruction has not been given to register any of the difference data, the reception unit 45 proceeds to step S9. If an instruction has been given to register any of the difference data, the registration unit 46 registers the instructed difference data in the database storage device 30 as teacher data (step S8), and proceeds to step S9.

In step S9, the registration unit 46 determines whether or not an instruction to end has been given (step S9). The user can instruct to end or continue by performing a predetermined operation using the operation unit 15. If an instruction to end has not been given, the registration unit 46 returns to step S5. If further difference data is to be registered, the user will instruct to continue without instructing to end. If an instruction to end has been given, the registration unit 46 ends the assistance process.

(4) Effects In the support device 40 according to this embodiment, abnormal data indicating an image of an inspection object previously determined to have a defect is acquired by the abnormal data acquisition unit 41. Normal data indicating an image of a normal inspection object is acquired by the normal data acquisition unit 42 so as to correspond to the abnormal data acquired by the abnormal data acquisition unit 41. Differential data indicating the difference between the abnormal data acquired by the abnormal data acquisition unit 41 and normal data acquired by the normal data acquisition unit 42 so as to correspond to the abnormal data is generated by the differential data generation unit 43. It is highly likely that the portion of the inspection object in the image displayed based on the differential data will require reinspection.

Therefore, based on the differential data generated by the differential data generation unit 43, an image of a portion of the object to be inspected is presented by the presentation unit 44. The selection of the differential data presented by the presentation unit 44 is accepted by the acceptance unit 45. The differential data selected and accepted by the acceptance unit 45 is registered as teacher data in the database storage device 30 by the registration unit 46. In this case, the user can select the differential data corresponding to the desired image while visually checking the presented image, thereby registering the selected differential data as teacher data. This makes it easier to create teacher data.

The receiving unit 45 further receives corrections to the difference data presented by the presenting unit 44. The registering unit 46 registers the corrected difference data as training data. In this case, the user can more appropriately correct the parts of the inspection object that require re-inspection in the image based on the difference data. This makes it possible to create training data for performing inspections with higher accuracy.

(5) Modifications Fig. 6 is a diagram showing various data used to create teacher data in the first modification. In the first modification, the user can register in advance a non-inspection area indicating an area outside the inspection area in the inspection device 20. When the non-inspection area is registered, the inspection device 20 sets the non-inspection area in the generated image data.

Therefore, as shown in the upper left of FIG. 6, a non-inspection area is set in the abnormal data acquired by the abnormal data acquisition unit 41. Similarly, as shown in the lower left of FIG. 6, a non-inspection area is set in the normal data acquired by the normal data acquisition unit 42. In this case, as shown in the right part of FIG. 6, the difference data generation unit 43 generates difference data by calculating the difference in each pixel value between the abnormal data and the normal data corresponding to the abnormal data, with the set non-inspection area excluded. The pixel values of the non-inspection area in the difference data may be set to 0.

In this case, it is possible to prevent parts outside the inspection area from being labeled as parts requiring reinspection. This makes it possible to create training data for more accurate inspection. In addition, because parts outside the inspection area are not reinspected, the created training data can be used to quickly reinspect the board.

Figure 7 is a diagram showing various data used to create teacher data in the second modified example. As shown in the upper left of Figure 7, the abnormality data acquisition unit 41 acquires defect information indicating the type of defect in the board in the image indicated by the abnormality data, along with the abnormality data. As shown in the right part of Figure 7, the difference data generation unit 43 adds the defect information acquired by the abnormality data acquisition unit 41 to the generated difference data. In the image display area 51 of the GUI 50, the image based on the difference data may be displayed in a manner (e.g., color) indicating the type of defect.

In this case, the user does not need to add defect information to the training data. This reduces the burden on the user and improves the work efficiency of creating training data. In addition, since there is no risk of user errors, more accurate training data can be created.

Figure 8 is a diagram showing various data used to create training data in the third modified example. In the third modified example, the user can set the inspection device 20 in advance to perform binarization processing on the image data. When it is set to perform binarization processing, the inspection device 20 generates binarized image data.

Therefore, as shown in the upper left of FIG. 8, the abnormal data acquisition unit 41 acquires binarized abnormal data. Similarly, as shown in the lower left of FIG. 8, the normal data acquisition unit 42 acquires binarized normal data. As shown in the right of FIG. 8, the differential data generation unit 43 generates differential data by calculating the difference in pixel values between the binarized abnormal data and the binarized normal data corresponding to the abnormal data.

In this case, the amount of image data is reduced, so training data can be created quickly. In addition, by using the created training data, the board can be re-inspected quickly.

[2] Second embodiment In the first embodiment, the normal data acquisition unit 42 acquires normal data so that one normal data corresponds to one abnormal data, but the embodiment is not limited to this. Below, the support process in the second to fourth embodiments will be described with respect to the differences from the support process in the first embodiment.

Figure 9 is a diagram showing various data used to create teacher data in the second embodiment. In this embodiment, as shown in the lower left of Figure 9, multiple normal data are acquired so that multiple normal data correspond to one abnormal data. The differential data generation unit 43 generates differential data by calculating the difference in each pixel value between one abnormal data and each normal data corresponding to the abnormal data. Therefore, as shown in the right part of Figure 9, multiple differential data are generated corresponding to one abnormal data.

With this configuration, a large amount of difference data is generated from one piece of abnormal data. This can improve the work efficiency of creating training data.

[3] Third embodiment Fig. 10 is a diagram showing various data used to create teacher data in the third embodiment. In this embodiment, as shown in the lower left of Fig. 10, a plurality of normal data are acquired so that a plurality of normal data correspond to one abnormal data. The differential data generating unit 43 generates differential data by calculating the difference in each pixel value between one abnormal data and the average of the plurality of normal data corresponding to the abnormal data. Therefore, as shown in the right part of Fig. 10, one differential data is generated corresponding to one abnormal data.

With this configuration, even if a noise component unrelated to a defect accidentally gets mixed into any of the multiple normal data, the multiple normal data are averaged, so the noise component has almost no effect on the pixel value of the normal data after averaging. Therefore, by using the normal data after averaging, it is possible to create training data for performing inspections with higher accuracy.

[4] Fourth embodiment Fig. 11 is a diagram showing various data used to create teacher data in the fourth embodiment. In this embodiment, as shown in the upper left of Fig. 11, normal data is acquired so that one normal data corresponds to a plurality of abnormal data. The differential data generating unit 43 generates differential data based on each abnormal data and the normal data corresponding to the abnormal data. Therefore, as shown in the right part of Fig. 11, a plurality of differential data corresponding to each of the plurality of abnormal data are generated.

With this configuration, since it is not necessary to acquire normal data each time abnormal data is acquired, teacher data can be created at high speed using common normal data. In this embodiment, image data showing an image of a normal board that has been prepared in advance may be acquired as normal data by the normal data acquisition unit 42. In this case, step S2 in the support process may be executed before step S1.

The image data prepared in advance may be any of the image data generated by the inspection device 20, but may also be master data such as CAD data showing a design drawing of the board. Here, depending on the processing accuracy of the board by etching or the like, the pattern width of the board in the image or the radius of curvature of the corners of the pattern will differ between the image data generated by the inspection device 20 and the master data. Therefore, the master data may be modified based on the processing accuracy of the board so that the pattern width or the radius of curvature of the corners of the pattern in the image is changed.

In the first to third embodiments, image data showing an image of a board that has not been determined in advance to have a defect by the inspection device 20 is acquired from the inspection device 20 as normal data, but the embodiments are not limited to this. At least one of the normal data may be master data prepared in advance or master data that has been corrected.

[5] Other embodiments In the above embodiment, the support device 40 includes the reception unit 45 and the registration unit 46, but the embodiment is not limited to this. The support device 40 does not need to include the reception unit 45 and the registration unit 46. Even in this case, the user can create teacher data for re-inspecting the part by checking the image presented on the GUI 50. This makes it possible to easily create teacher data for performing inspections with high accuracy.

[6] Correspondence between each component of the claims and each part of the embodiment Below, examples of correspondence between each component of the claims and each element of the embodiment will be described, but the present invention is not limited to the following examples. Various other elements having the configuration or function described in the claims can also be used as each component of the claims.

In the above embodiment, the support device 40 is an example of a teacher data creation support device, the abnormal data acquisition unit 41 is an example of an abnormal data acquisition unit, and the normal data acquisition unit 42 is an example of a normal data acquisition unit. The differential data generation unit 43 is an example of a differential data generation unit, the presentation unit 44 is an example of a presentation unit, the reception unit 45 is an example of a reception unit, and the registration unit 46 is an example of a registration unit.
[7] Reference form
(1) A teacher data creation support device in a first reference form is a teacher data creation support device that supports the creation of teacher data to be used for re-inspection of an inspection object, and includes an abnormal data acquisition unit that acquires abnormal data indicating an image of an inspection object that has been determined in advance to have a defect, a normal data acquisition unit that acquires normal data indicating an image of a normal inspection object corresponding to the abnormal data acquired by the abnormal data acquisition unit, a differential data generation unit that generates differential data indicating the difference between the abnormal data acquired by the abnormal data acquisition unit and the normal data acquired by the normal data acquisition unit so as to correspond to the abnormal data, and a presentation unit that presents an image of a portion of the inspection object based on the differential data generated by the differential data generation unit.
In this teacher data creation support device, an image of a portion of an inspection object is presented based on difference data indicating the difference between abnormal data indicating an image of an inspection object previously determined to have a defect and normal data indicating an image of a normal inspection object. The portion of the inspection object in the image shown based on the difference data is likely to require reinspection. Therefore, by checking the presented image, a user can create teacher data for reinspecting that portion. This makes it possible to easily create teacher data for performing inspections with high accuracy.
(2) The teacher data creation support device may further include a reception unit that receives a selection of the difference data presented by the presentation unit, and a registration unit that registers the difference data selected by the reception unit as teacher data. In this case, the difference data selected by the user is registered as teacher data. This makes it easier to create teacher data.
(3) The receiving unit may further receive corrections to the difference data presented by the presenting unit, and the registering unit may register the corrected difference data as training data. In this case, it is possible to more appropriately correct a portion of the inspection object that requires re-inspection in the image based on the difference data. This makes it possible to create training data for performing inspections with higher accuracy.
(4) The normal data acquisition unit may acquire a plurality of normal data corresponding to the abnormal data, and the differential data generation unit may generate a plurality of differential data based on the abnormal data and the plurality of normal data corresponding to the abnormal data. In this case, a large number of differential data are generated from one abnormal data. This can improve the work efficiency of creating teacher data.
(5) The normal data acquisition unit may acquire a plurality of normal data corresponding to the abnormal data, and the differential data generation unit may generate differential data based on the abnormal data and an average of the plurality of normal data corresponding to the abnormal data. With this configuration, even if a noise component unrelated to a defect accidentally gets mixed into any of the plurality of normal data, the plurality of normal data are averaged, so that the noise component has almost no effect on the pixel value of the normal data after averaging. Therefore, it is possible to create teacher data for performing inspection with higher accuracy.
(6) The normal data acquisition unit may acquire normal data corresponding to a plurality of abnormal data, and the differential data generation unit may generate differential data based on each abnormal data and the normal data corresponding to the abnormal data. In this case, teacher data can be created at high speed using the common normal data.
(7) The normal data acquired by the normal data acquisition unit may include image data representing an image of an inspection object that has not been determined to have a defect in advance. In this case, the normal data representing an image of a normal inspection object can be easily acquired.
(8) The normal data acquired by the normal data acquisition unit may include master data representing a design drawing of the inspection object. In this case, the normal data representing a normal image of the inspection object can be easily acquired.
(9) The normal data acquisition unit may acquire master data that has been corrected based on the processing accuracy of the object to be inspected as the normal data. With this configuration, even if the area to be inspected is minute, normal data that shows an image of a normal object to be inspected can be easily acquired.
(10) A non-inspection target area may be set in the abnormal data acquired by the abnormal data acquisition unit and in the normal data acquired by the normal data acquisition unit, and the difference data generation unit may generate the difference data by excluding the set non-inspection target area. In this case, the image represented by the difference data is prevented from including a portion outside the inspection target area. This makes it possible to create teacher data for performing inspections with higher accuracy.
(11) The abnormality data acquisition unit may further acquire defect information indicating a type of defect in the inspection object corresponding to the acquired abnormality data, and the differential data generation unit may add the defect information acquired by the abnormality data acquisition unit to the generated differential data. In this case, the user does not need to perform the task of adding the defect information to the teacher data. This reduces the burden on the user and improves the work efficiency of creating the teacher data. In addition, since there is no error associated with the user's work, more accurate teacher data can be created.
(12) The abnormal data acquisition unit may acquire binarized abnormal data, and the normal data acquisition unit may acquire binarized normal data. In this case, the amount of abnormal data and normal data is reduced, so that teacher data can be created at high speed.
(13) A teacher data creation support method according to a second reference form is a teacher data creation support method that supports the creation of teacher data to be used for re-inspection of an inspection object, and includes the steps of acquiring abnormal data indicating an image of an inspection object that has been determined in advance to have a defect, acquiring normal data indicating an image of a normal inspection object that corresponds to the acquired abnormal data, generating differential data indicating the difference between the acquired abnormal data and normal data acquired to correspond to the abnormal data, and presenting an image of a portion of the inspection object based on the generated differential data.
According to this method for supporting the creation of training data, an image of a portion of an inspection object is presented based on differential data indicating the difference between abnormal data indicating an image of an inspection object previously determined to have a defect and normal data indicating an image of a normal inspection object. The portion of the inspection object in the image shown based on the differential data is likely to require reinspection. Therefore, by checking the presented image, a user can create training data for reinspecting that portion. This makes it possible to easily create training data for performing inspections with high accuracy.

10... Processing device, 11... CPU, 12... RAM, 13... ROM, 14... Storage device, 15... Operation unit, 16... Display device, 17... Input/output I/F, 18... Bus, 20... Inspection device, 30... Database storage device, 40... Support device, 41... Abnormal data acquisition unit, 42... Normal data acquisition unit, 43... Differential data generation unit, 44... Presentation unit, 45... Reception unit, 46... Registration unit, 50... GUI, 51... Image display area, 52... Registration button, 53... Correction button, 100... Processing system

Claims (12)

A teacher data creation support device that supports the creation of teacher data used for re-inspection of an inspection object,
an abnormality data acquisition unit that acquires abnormality data indicating an image of an inspection object that has been determined in advance to have a defect;
a normal data acquisition unit that acquires normal data representing an image of a normal inspection object corresponding to the abnormal data acquired by the abnormal data acquisition unit;
a differential data generating unit that generates differential data indicating a difference between the abnormal data acquired by the abnormal data acquiring unit and normal data acquired by the normal data acquiring unit so as to correspond to the abnormal data;
a presentation unit that presents an image of a portion of an object to be inspected based on the difference data generated by the difference data generation unit;
a reception unit that receives a selection of the difference data presented by the presentation unit;
A teacher data creation support device comprising: a registration unit that registers the difference data selected and accepted by the acceptance unit as teacher data. The accepting unit further accepts a correction to the difference data presented by the presenting unit;
2. The teacher data creation support device according to claim 1, wherein the registration unit registers the corrected difference data as teacher data. the normal data acquisition unit acquires a plurality of normal data corresponding to the abnormal data;
3. The teacher data creation support device according to claim 1, wherein the difference data generating unit generates a plurality of difference data based on the abnormal data and a plurality of normal data corresponding to the abnormal data. the normal data acquisition unit acquires a plurality of normal data corresponding to the abnormal data;
3. The teacher data creation support device according to claim 1, wherein the difference data generating unit generates the difference data based on the abnormal data and an average of a plurality of normal data corresponding to the abnormal data. the normal data acquisition unit acquires normal data corresponding to a plurality of abnormal data;
3. The teacher data creation support device according to claim 1, wherein the differential data generating unit generates differential data based on each abnormal data and normal data corresponding to the abnormal data. The teacher data creation support device according to any one of claims 1 to 5, wherein the normal data acquired by the normal data acquisition unit includes image data showing an image of an inspection object that has not been determined to have a defect in advance. The teacher data creation support device according to any one of claims 1 to 6, wherein the normal data acquired by the normal data acquisition unit includes master data showing a design drawing of an object to be inspected. The teaching data creation support device according to claim 7, wherein the normal data acquisition unit acquires master data that has been corrected based on the processing accuracy of the object to be inspected as normal data. a non-inspection target area is set in the abnormal data acquired by the abnormal data acquisition unit and in the normal data acquired by the normal data acquisition unit;
The teacher data creation support device according to claim 1 , wherein the difference data generation unit generates difference data by excluding a set non-inspection target area. the abnormality data acquisition unit further acquires defect information indicating a type of defect in the inspection object corresponding to the acquired abnormality data;
The teacher data creation support device according to any one of claims 1 to 9, wherein the difference data generation unit adds defect information acquired by the abnormality data acquisition unit to the generated difference data. The abnormal data acquisition unit acquires the binarized abnormal data,
The teacher data creation support device according to claim 1 , wherein the normal data acquisition unit acquires normal data that has been binarized. A method for supporting the creation of training data used for re-inspection of an inspection object, comprising:
The teacher data creation support device
Obtaining abnormality data indicating images of inspection objects that have been previously determined to have defects;
acquiring normal data representing an image of a normal inspection object corresponding to the acquired abnormal data;
generating difference data indicating a difference between the acquired abnormal data and normal data acquired so as to correspond to the abnormal data;
presenting an image of a portion of the object to be inspected based on the generated difference data;
Accept the presented differential data selection,
A method for supporting the creation of teacher data, in which differential data whose selection has been accepted is registered as teacher data.

Images