JP2013228232A - Substrate appearance inspection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable parallel execution of an inspection by image processing and a visual inspection by an inspector so as to reduce inspection time as a whole.SOLUTION: A substrate appearance inspection method includes: an imaging step of sequentially imaging an inspection target substrate at each area by a camera; an image recognition step of recognizing an image of a defect candidate in the inspection target substrate by sequentially processing the image at each area obtained at the imaging step; and a defect candidate area image display step to be executed before a completion of image recognition processing of the last area at the image recognition at the latest, the step of displaying an image of an area in which a defect candidate is recognized at the image recognition step, being an image of an area having undergone image the recognition processing before the last area, so as to allow a visual inspection by a user.

Description

  The present invention relates to a substrate visual inspection method using a camera.

  Conventionally, this kind of substrate appearance inspection method is known (see, for example, Patent Document 1).

Japanese Patent No. 4436523

  By the way, in a configuration in which a substrate to be inspected is imaged with a camera and a defect candidate is recognized by image processing, generally, a visual inspection by an inspector is performed after the defect candidate is image-recognized. However, when it is necessary to capture a plurality of images with respect to one substrate to be inspected, for example, when the substrate to be inspected is relatively large, the inspection is performed until the image recognition processing for all the images is completed. The visual inspection by the person cannot be started, and there is a problem that the inspection time as a whole becomes long.

  SUMMARY OF THE INVENTION An object of the present invention is to provide a substrate appearance inspection method capable of performing both inspection by image processing and visual inspection by an inspector in parallel, and shortening the entire inspection time.

In order to achieve the above object, according to one aspect of the present invention, an imaging step of sequentially imaging a substrate to be inspected by a camera for each area;
An image recognition step for sequentially processing images for each area obtained by the imaging step, and recognizing a defect candidate on the substrate to be inspected,
The image of the area in which the defect candidate is recognized by the image recognition step so that it is executed at the latest before the image recognition processing of the last area by the image recognition step is completed and visual inspection by the user is possible, And a defect candidate area image display step for displaying an image of an area that has undergone image recognition processing before the last area.

  According to the present invention, it is possible to perform both an inspection by image processing and a visual inspection by an inspector in parallel, and a substrate appearance inspection method capable of shortening the entire inspection time can be obtained.

It is a flowchart which shows an example of the board | substrate external appearance inspection method by this invention. It is a block diagram which shows an example of the board | substrate external appearance inspection system 1 which performs the board | substrate external appearance inspection method shown in FIG. It is a top view which shows roughly an example of each area of a to-be-inspected board | substrate.

  The best mode for carrying out the present invention will be described below with reference to the drawings.

  FIG. 1 is a flowchart showing an example of a substrate appearance inspection method according to the present invention. FIG. 2 is a configuration diagram illustrating an example of the substrate appearance inspection system 1 that executes the substrate appearance inspection method illustrated in FIG. 1. FIG. 3 is a plan view schematically showing an example of each area of the substrate to be inspected. The substrate appearance inspection method is executed in an inspection process in the manufacturing process in which the substrate to be inspected is a product.

  The substrate to be inspected may be any substrate having a finite outer shape, such as glass, an organic EL panel, a film, and the like. Alternatively, the substrate to be inspected may be a substrate that is substantially infinitely continuous. Moreover, in the case of glass, glass itself may be sufficient and glass with a film | membrane may be sufficient. In the latter case, the presence or absence of glass defects and / or the presence or absence of film defects may be inspected in the inspection step. The film may have an arbitrary function, for example, a film having a heat insulating function. The size of the substrate to be inspected may be arbitrary, but is large enough that the entire area cannot be imaged with one shot of the camera 10 as described later (that is, it is necessary to divide the image into a plurality of areas and image each area). The size is preferred. Further, the substrate to be inspected may be a transparent substrate such as glass or an opaque substrate.

  Here, first, the board appearance inspection system 1 will be described with reference to FIG.

  The board appearance inspection system 1 includes a camera 10, an illumination device 12, an XY biaxial stage 14, and a control device 18.

  The camera 10 may be an area sensor in which CCD elements are arranged in the vertical and horizontal directions, or may be a line sensor in which CCD elements are arranged in a horizontal row. The area sensor can capture an image two-dimensionally, and the line sensor can capture a one-dimensional image. In addition to the CCD, a CMOS or the like may be used as the imaging element. Hereinafter, the case where the camera 10 is an area sensor will be described. The orientation of the camera 10 is determined according to the defect to be detected, the type of the substrate to be inspected, and the like, and may be arranged so that the optical axis faces the direction perpendicular to the substrate to be inspected. On the other hand, the optical axis may be arranged in an oblique direction. The camera 10 may be arranged at a position for imaging the front surface of the substrate to be inspected, may be arranged at a position for imaging the back surface of the substrate to be inspected, or images the front and back surfaces of the substrate to be inspected. Two or more units may be arranged.

  The illuminating device 12 is arranged in an appropriate positional relationship with respect to the camera 10 so that a defect in the inspected substrate can be recognized. The arrangement / orientation / number of the illuminating devices 12 can vary depending on the defect to be detected, the type of the substrate to be inspected, the position and orientation of the camera 10, and the like. For example, the illumination device 12 may be disposed on the same side as the camera 10 with respect to the substrate to be inspected, or may be disposed on the opposite side of the camera 10 with respect to the substrate to be inspected.

  The XY biaxial stage 14 is a stage on which a substrate to be inspected is placed, and is movable along two orthogonal axes (XY axes). The XY biaxial stage 14 typically moves with two axes in a horizontal plane. In this case, the stage of the XY biaxial stage 14 (support surface of the substrate to be inspected) provides a horizontal plane.

  The control device 18 includes a computer and controls the camera 10, the illumination device 12, and the XY biaxial stage 14. For example, the control device 18 outputs an imaging command or the like to the camera 10. Further, the control device 18 outputs a stage movement command or the like to the XY biaxial stage 14. The stage movement command is output in cooperation with the imaging command.

  In this embodiment, the camera 10 and the XY biaxial stage 14 are controlled so that the substrate to be inspected is imaged by the camera 10 for each area. In the example shown in FIG. 3, the inspected substrate 70 is divided into 20 areas. The area corresponds to the area of the substrate to be inspected that is imaged by the camera 10 in one imaging. Accordingly, the size and shape of the area may differ depending on the pixel configuration of the camera 10 and the positional relationship such as the vertical distance between the substrate 70 to be inspected and the camera 10.

  In the example illustrated in FIG. 3, for example, the camera 10 and the XY biaxial stage 14 may be controlled so that each area is sequentially imaged in ascending order from the first area to the twentieth area. In this case, the camera 10 first images the first area, and then the inspected substrate 70 is moved in the Y direction (leftward) by the XY biaxial stage 14, and the camera 10 images the second area. Similarly, when the inspected substrate 70 is moved in the Y direction by the XY biaxial stage 14 and the third, fourth, and fifth areas are sequentially imaged, the inspected substrate 70 is moved in the X direction by the XY biaxial stage 14, and the first Six areas are imaged. Next, the inspected substrate 70 is moved in the Y direction (rightward) by the XY biaxial stage 14, and the camera 10 images the seventh area. In this way, the substrate to be inspected is imaged by the camera 10 for each area.

  In the above description, the positional relationship between the camera 10 and the substrate to be inspected 70 is changed by fixing the camera 10 and moving the substrate 70 to be inspected by the XY biaxial stage 14. It may be moved, or both the camera 10 and the inspected substrate 70 may be moved. Moreover, the imaging order of each area is also arbitrary. Moreover, although each area shown in FIG. 3 is mutually independent, a part may overlap with an adjacent area.

  The board appearance inspection system 1 further includes an image processing device 20, a display device 22, a visual result input device 24, and an inspection data storage device 26.

  The image processing device 20 is composed of a computer. The image processing apparatus 20 processes an image obtained from the camera 10 and recognizes an image of a defect that may exist on the inspected substrate. The defect of the image recognition target is arbitrary, but is typically a defect that causes a problem in the quality of the substrate to be inspected. For example, in the case of glass with a film, it may be a film defect (for example, a bubble or a foreign material in the film) or a glass defect (for example, a foreign material in the glass). Further, in the case of a laminated substrate, it may be a deviation in lamination, adhesion of foreign matter (dirt), defect, or the like. As the defect image recognition method, any method may be used according to the type (feature) of the defect. For example, the defect image recognition method may be a method of recognizing an image of a defect (abnormal part) by comparing with reference image data obtained by imaging a substrate to be inspected without a defect. Note that the defect recognized by the image processing apparatus 20 is finally judged (inspected) whether it is a defect by human eyes. Therefore, hereinafter, the defect recognized by the image processing apparatus 20 is referred to as a “defect candidate”.

  The image processing device 20 sequentially executes image recognition processing for each image of each area obtained from the camera 10. The defect candidate recognition result (the presence / absence of a defect candidate, the position of the defect candidate, etc.) by the image processing apparatus 20 is obtained for each image in each area. The defect candidate recognition result may be sequentially stored in a predetermined storage device together with the image of each area.

  The display device 22 is a display device for displaying an image to the examiner. The display device 22 may be any display device, for example, a liquid crystal display. The display device 22 may be a display device having only a display function, or may be a display device including a control device (computer) inside. When the display device has only the display function, the display state of the display device 22 may be controlled by the visual result input device 24, for example. In the following description, it is assumed that the display state of the display device 22 is controlled by the visual result input device 24.

  The display device 22 displays an image (hereinafter, also referred to as “defect candidate area image”) that is an image obtained from the camera 10 and related to an area in which the defect candidate is image-recognized by the image processing device 20. The display device 22 may display a defect candidate area image in response to a request from a user (typically, an inspector), or visually when a defect candidate is recognized by the image processing device 20. Upon receiving a command from the result input device 24 (automatically), a defect candidate area image related to the defect candidate may be displayed. The display device 22 may have a function of enlarging and displaying the defect candidate area image in response to a request from the user. The defect candidate area image displayed on the display device 22 may be the image itself obtained from the camera 10, or a mark for informing the user of the position of the defect candidate on the image obtained from the camera 10. May be an image in which is superimposed.

  The visual result input device 24 is composed of a computer. A user interface 25 is connected to the visual result input device 24. The user interface 25 may be any user interface 25 and may be, for example, a mouse, a keyboard, a touch panel display, a voice recognition device, or the like. The connection form between the visual result input device 24 and the user interface 25 is arbitrary, and may be wireless communication, wired communication, or a combination thereof.

  The visual result input device 24 processes the visual inspection result by the inspector input via the user interface 25. Note that the user interface 25 of the visual result input device 24 is typically disposed near the display device 22. The inspector visually checks the defect candidate area image displayed on the display device 22, determines whether the defect candidate displayed there is a defect, and determines the determination result (visual inspection result) by the user. Input from the interface 25. The visual result input device 24 inspects the visual inspection results (for example, two types of defects (NG) and non-defective (OK), or three types of defects, non-defective and pending) related to the input defect candidate area image. Store in the storage device 26.

  As an example, in this example, the visual result input device 24 controls the display state of the display device 22. In this case, an image (including a defect candidate area image) obtained from the camera 10 may be stored in a storage device accessible by the visual result input device 24. For example, each image from the camera 10 that is sequentially supplied via the image processing device 20 may be sequentially stored in the main storage device inside the visual result input device 24. In this case, the visual recognition result input device 24 may be supplied with the defect candidate recognition result from the image processing device 20 together with the image of each area as needed.

  The inspection data storage device 26 is an auxiliary storage device, and is composed of, for example, an HDD (Hard Disk Drive). The inspection data storage device 26 may be an auxiliary storage device built in the visual result input device 24 or may be an accessible external auxiliary storage device. The inspection data storage device 26 stores the visual inspection result input via the visual result input device 24. The visual inspection result stored in the inspection data storage device 26 is used in a process subsequent to the inspection process. For example, the process after the inspection process may be a process of performing maintenance such as repair of the substrate or discarding the substrate according to the inspection result of the inspection process. The inspection data storage device 26 may store defect position information (information indicating which position on the substrate) in addition to the visual inspection result. The defect position information may be expressed by coordinates in a predetermined coordinate system defined for the substrate, for example. The defect position information may be information input together with the visual inspection result by the inspector, or position information calculated by the image processing apparatus 20 (position information of defect candidates at the time of calculation). There may be. Thereby, in the process after the inspection process, for example, when performing repair, it is possible to easily identify the corresponding defect. Maintenance such as repair may be manual work by an operator, automatic work by a robot, or a combination thereof. Further, in addition to the visual inspection result, the inspection data storage device 26 may store additional information indicating the type and degree of the defect. In addition to the visual inspection result, the corresponding defect candidate area image may be stored in the inspection data storage device 26.

  Next, an example of a substrate appearance inspection method realized by the above-described substrate appearance inspection system 1 shown in FIG. 2 will be described with reference to FIG.

  In FIG. 1, the operations by the visual inspector are shown in the same time series together with the operations of the camera 10, the image processing device 20, the display device 22, and the inspection data storage device 26. FIG. 1 shows a timing chart of a substrate appearance inspection method executed for a certain substrate to be inspected.

  First, the camera 10 starts its operation from time t0 and sequentially images each area of the substrate to be inspected. In the example illustrated in FIG. 1, the camera 10 sequentially captures images from the first area to the Nth area, and ends operation at time t1 when the Nth area has been captured. The operating time (= t1-t0) of the camera 10 depends on the number of areas (N), the imaging time of the camera 10 per shot, the operating time of the XY 2-axis stage 14, and the like.

  After time t0, the image processing apparatus 20 starts to operate upon receiving an image of the first area from the camera 10. The image processing apparatus 20 sequentially processes images of each area of the board to be inspected, which are sequentially captured by the camera 10. In the example illustrated in FIG. 1, the image processing apparatus 20 starts processing the first area image before the camera 10 captures the second area image. However, the process (defect candidate recognition process) of the image processing apparatus 20 may be executed at any timing as long as it is sequentially executed in parallel with the imaging of the camera 10. Note that typically, the image processing apparatus 20 processes the images of each area in the order synchronized with the order of acquisition of images by the camera 10 (in this example, the ascending order from the first area to the Nth area). In this case, the operation of the image processing device 20 ends when the image of the Nth area has been processed (time t2 in the example shown in FIG. 1).

  As shown in FIG. 1, the defect candidate recognition process by the image processing apparatus 20 is performed by the process of the visual result input device 24 (for example, the storage process of the visual result from the visual inspector in the inspection data storage device 26, or the display The defect candidate area image display process on the apparatus 22 is executed in parallel. That is, since the image processing device 20 and the inspection data storage device 26 are configured by separate computers, the image processing device 20 is not affected by the processing of the inspection data storage device 26, and is a defect candidate for the image in each area. Recognition processing can be continuously executed sequentially.

  The display device 22 displays an image (defect candidate area image) related to an area in which the defect candidate is recognized by the image processing device 20 at a predetermined timing after time t0. The defect candidate area image may be automatically and sequentially displayed every time the defect candidate is image-recognized by the image processing device 20, or when a request from a user (typically an inspector) is input. May be displayed in response to the request. In the latter case, when a new defect candidate area image is generated, the display device 22 displays that fact, and when a display request is input from the user in response thereto, the defect candidate area image is displayed. It is good to do. The request from the user is processed by a control device (in this example, the visual result input device 24) that controls the display of the display device 22. For example, a request from a user is input via the user interface 25, and in response to this, the visual result input device 24 may display a defect candidate area image on the display device 22. In the example shown in FIG. 1, the defect candidate is recognized in the image of the second area. Upon receiving the defect candidate, the display device 22 immediately after the defect candidate is recognized by the image processing device 20, the defect related to the defect candidate. A candidate area image is displayed. The display device 22 typically displays all defect candidate area images, and substantially ends the operation when a visual inspection result corresponding to the defect candidate area image is input.

  In the example shown in FIG. 1, the display device 22 displays a plurality of defect candidate area images for one substrate to be inspected. In this case, the switching of the defect candidate area image may be executed in response to a switching request from the user. For example, the visual result input device 24 may display the next defect candidate area image on the display device 22 when a switching request is input from the user via the user interface 25.

  The visual inspector visually inspects the defect candidate area image displayed on the display device 22. That is, the visual inspector views the defect candidate area image displayed on the display device 22 to determine whether or not the defect candidate is a defect, and inputs the determination result (visual inspection result) through the user interface 25. To do. The visual inspection may involve display switching such as an enlarged display of the defect candidate area image on the display device 22. The visual inspection by the visual inspector is executed for each defect candidate area image displayed on the display device 22. The visual inspection by the visual inspector is executed in synchronization with the display of the defect candidate area image on the display device 22. When a defect candidate area image is displayed on the display device 22, a state in which a visual inspection result related to the defect candidate area image can be input (a state waiting for input of a visual inspection result) may be formed. . Actually, the visual inspection result is input after a certain amount of time (time required for visual inspection) has elapsed since the defect candidate area image was displayed on the display device 22. In the example shown in FIG. 1, the visual inspection result input for the first defect candidate area image and the visual inspection result input for the second defect candidate area image are performed in the last area (Nth) by the image processing apparatus 20. This is executed before the image processing of (Area) is completed. On the other hand, the visual inspection result input for the last defect candidate area image is executed after the image processing of the last area (Nth area) by the image processing apparatus 20 is completed (that is, after time t2).

  The inspection data storage device 26 stores the visual inspection result input by the visual inspector as needed. That is, the visual result input device 24 stores the input visual inspection result in the inspection data storage device 26 every time the visual inspection result is input by the visual inspector. The visual result input device 24 stores, in the main storage device, visual inspection results related to a plurality of defect candidate area images input by a visual inspector (however, visual inspection results related to the same substrate to be inspected). When the visual inspection result relating to the last defect candidate area image is input, it may be stored in the inspection data storage device 26 at once. In any case, at the time when the inspection data storage device 26 stores the visual inspection result related to the last defect candidate area image (time t3 in FIG. 1), the substrate appearance inspection related to one specific inspection substrate is performed. Completed. In the example shown in FIG. 1, three visual inspection results are stored corresponding to three defect candidate area images for one specific substrate to be inspected. The number of can vary. For example, there may be no defect candidate area image for one substrate to be inspected.

  As described above, the substrate appearance inspection shown in FIG. 1 substantially starts from the time (time t0) when the camera 10 captures an image of the first area for a specific one substrate to be inspected, and stores inspection data. The process ends when the device 26 stores the visual inspection result relating to the last defect candidate area image (time t3).

  Here, according to the substrate appearance inspection method shown in FIG. 1, as described above, as soon as the defect candidate area image is obtained, the defect candidate area image can be displayed on the display device 22. Thereby, the visual inspection by the inspector can be started before the image processing of the last area (the Nth area in this example) by the image processing apparatus 20 is completed. For example, in the example illustrated in FIG. 1, an area before the Nth area (for example, a second area) on the display device 22 before time t2 when the image processing of the Nth area by the image processing apparatus 20 ends. The defect candidate area image concerning is displayed. Thus, after all the defect candidate recognition processing from the image of the first area to the image of the Nth area, the defect candidate area image related to the area before the Nth area is displayed earlier than when displaying the defect candidate area image. Visual inspection by the inspector can be started. As a result, it is necessary to divide the image into a large number of areas, and a defect caused by the image processing apparatus 20 is required even for a large inspected substrate that requires a relatively long time to shoot all the images. The candidate recognition process and the visual inspection by the inspector can be executed in cooperation, and as a result, the inspection time required as a whole can be shortened.

  In the example shown in FIG. 1, the visual inspection result input for the first defect candidate area image and the visual inspection result input for the second defect candidate area image are performed in the last area (Nth) by the image processing apparatus 20. It is executed before the time (time t2) when the image processing of (area) ends. However, depending on the number of areas N, the image processing time, the visual inspection time, the appearance of the defect candidate area image, etc., the visual inspection result input for the defect candidate area image is performed by the image processing of the last area by the image processing device 20. It may not complete before the point of termination. For example, in the example shown in FIG. 1, when N = 2 and the first defect candidate area image is the image of the first area, the visual inspection result for the image of the first area is the last area (second There may be a case where the input cannot be performed before the time point when the image processing of (Area) ends. Even in this case, according to the present embodiment, the visual inspection for the image of the first area can be started before the end of the image processing of the last area (second area) ( That is, since the defect candidate recognition process for the first area can be started immediately after completion, the visual inspection for the image of the first area is started from the point when the defect candidate recognition process for the last area (second area) is completed. Compared to the configuration, the visual inspection can be started earlier, and the inspection time can be shortened accordingly.

  The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the above-described embodiments, and various modifications and substitutions can be made to the above-described embodiments without departing from the scope of the present invention. Can be added.

  For example, in the above-described embodiment, the image processing device 20 and the visual result input device 24 are realized by separate computers, but may be realized by a single computer. In this case, the single computer is a multi-core processor, and each CPU core realizes the functions of the image processing device 20 and the visual result input device 24. Further, the function of the image processing device 20 may be realized by a plurality of computers, and the function of the visual result input device 24 may be realized by a plurality of computers.

  In the above-described embodiment, the camera 10 has finished operating when the image of the Nth area is acquired. For example, in response to a request from the user via the user interface 25, the image of the specified area is displayed. May be operated to image again. In this case, when the visual inspection related to the last defect candidate area image is completed, by accepting such re-imaging, when all the visual inspection is completed, only a specific area is imaged again and re-inspected. In comparison, the inspection time can be shortened as a whole. In this case, the image processing apparatus 20 may perform the defect candidate recognition process again for the re-captured area. In the configuration in which the lens of the camera 10 can be switched (that is, the configuration in which the camera 10 is equipped with a variable magnification lens), the lens of the camera 10 has a higher magnification so that the defect candidate can be enlarged and observed. The image of the designated area may be taken again after switching. Alternatively, the camera 10 may be a main camera, a camera different from the camera 10 may be used as a sub camera, and an image of a specified area may be captured again by the sub camera. In this case, the sub camera may be provided with a lens having a higher magnification than that of the main camera (camera 10) or may be provided with a variable magnification so that the defect candidate can be enlarged and observed. According to such a configuration, during the imaging operation of the camera 10 (that is, before imaging of the Nth area is completed), the imaging operation (imaging of the image of the designated area) by the sub camera can be started. Only the inspection time can be shortened. In this case, the secondary camera may be placed on a planar motor, and the planar motion of the secondary camera (movement to an imaging position in a specific area) may be realized by driving the planar motor.

  In the above-described embodiment, maintenance such as repair is performed after completion of the inspection process (after visual inspection is completed for all defect candidates), but repair is performed during the inspection process. May be. That is, the maintenance work according to the visual inspection result input in the inspection process may be performed during the inspection process. The maintenance work may be performed manually, a robot work, or a combination thereof. In the case of work by a robot, the maintenance robot may be controlled in a manner that does not interfere with the imaging operation of the camera 10. In this case, maintenance work can be started during the imaging operation by the camera 10 (that is, before imaging of the Nth area is completed).

  In the embodiment described above, the control device 18 and the visual result input device 24 are realized by separate computers. However, the function of the control device 18 may be realized by a computer of the visual result input device 24.

  In the above-described embodiment, the camera 10 is an area sensor, but may be a line sensor as described above. In this case, each area corresponds to a range that can be imaged by a single imaging by the line sensor. Similarly, the image processing apparatus 10 may sequentially process images sequentially captured by the line sensor. Similarly, the display device 22 may display a defect candidate area image. However, in this case, in consideration of the visibility (ease of inspection) by the inspector, an image area including the defect candidate recognized by the image processing apparatus 10 may be enlarged and displayed in advance. The image of the area (pixels captured before that) may be combined and displayed.

  In the embodiment described above, the image processing apparatus 20 displays the defect candidate area image, but may display an image of an area in which no defect candidate is detected as necessary. For example, the image processing apparatus 20 may be configured to sequentially display an image of each area each time an image of each area is obtained from the camera 10. However, it is desirable that the inspector can easily call the defect candidate area image on the image processing apparatus 20 so that the inspector can easily perform visual inspection.

DESCRIPTION OF SYMBOLS 10 Camera 12 Illuminating device 14 XY 2 axis stage 18 Control apparatus 20 Image processing apparatus 22 Display apparatus 24 Visual result input apparatus 25 User interface 26 Inspection data storage apparatus 70 Board to be inspected

Claims (4)

An imaging step of sequentially imaging a substrate to be inspected by a camera for each area;
An image recognition step for sequentially processing images for each area obtained by the imaging step, and recognizing a defect candidate on the substrate to be inspected,
The image of the area in which the defect candidate is recognized by the image recognition step so that it is executed at the latest before the image recognition processing of the last area by the image recognition step is completed and visual inspection by the user is possible, A defect appearance area image display step for displaying an image of an area that has undergone image recognition processing before the last area.   The substrate appearance inspection method according to claim 1, wherein the defect candidate image display step includes forming a waiting state for inputting a visual inspection result for the displayed image of the area.   The user starts at the latest before the image recognition process for the last area in the image recognition step is completed, and the user determines the presence or absence of a defect in the inspection substrate based on the image displayed in the defect candidate area image display step. The substrate visual inspection method according to claim 1, further comprising a visual inspection step.   Visual inspection result recording that is executed before the image recognition processing of the last area by the image recognition step is completed at the latest and records the visual inspection result by the user for the defect candidate related to the image displayed by the defect candidate area image display step. The substrate visual inspection method according to claim 1, further comprising a step.

Images