Device and method for detecting panel line segment defects

Abstract

The present invention provides a device and method for detecting line segment defects of a panel. The device includes an image capture device, a sample storage device, and a processing device, wherein the processing device receives an initial image obtained by the image capture device from the area where the panel to be tested is placed, obtains the image of the panel to be tested corresponding to the panel to be tested from the initial image, obtains sample data from the sample storage device to set the image at the position of the sample line segment on the image of the panel to be tested corresponding to the sample line segment specified by the sample data as the image of the line segment to be tested, and checks the image of the line segment to be tested by executing a neural network classification program to determine whether the line segment to be tested has a line segment defect. In this way, panel quality problems caused by different human inspection standards can be reduced.

Description

Technical Field

[0001] The invention relates to a panel detection technology, and in particular to a device and method for detecting panel line segment defects. Background Art

[0002] In the modern society where large-size panels are very popular, how to ensure the quality of panels has become a very important part of the panel production process. In the existing technology, many manufacturers use the method of manually inspecting panels to confirm the manufacturing quality of panels. However, the method of manually inspecting panels is not only time-consuming and labor-intensive, but also because of the subjective factors of each person, it is very likely that there will be different inspection standards and ultimately cause the undesirable result of fluctuating panel quality. Summary of the invention

[0003] In view of this, an object of the present invention is to provide an apparatus and method for detecting panel line segment defects, which can automatically locate the position of the panel line segment, and can further send the image of the located panel line segment to a neural network classification program so that a pre-trained neural network classification program can determine whether there is a defect in the panel line segment, thereby reducing panel quality problems caused by different manual inspection standards.

[0004] From one perspective, the description of the present invention provides a device for detecting line segment defects of a panel, which is suitable for confirming whether there are line segment defects on the panel to be tested. This device includes an image capture device, a sample storage device and a processing device. The image capture device captures the area where the aforementioned panel to be tested is placed to obtain a corresponding initial image. The sample storage device stores sample data including the position of the sample line segment. The processing device is electrically connected to the aforementioned image capture device and the sample storage device, wherein the processing device receives the initial image from the image capture device and obtains the image of the panel to be tested corresponding to the panel to be tested from the initial image; in addition, the processing device also obtains sample data from the sample storage device to set the image corresponding to the sample line segment position on the image of the panel to be tested as the image of the line segment to be tested; and the processing device also executes a neural network classification program to check the image of the line segment to be tested and determine whether there are line segment defects in the line segment to be tested.

[0005] In one embodiment, the above-mentioned processing device divides the image block containing the line segment to be tested in the image of the panel to be tested into multiple image sub-blocks, and uses each image sub-block as input data of the neural network classification program so that the neural network classification program determines whether there is a line segment defect in the part of the line segment to be tested contained in the input image sub-block.

[0006] In one embodiment, when the neural network classification program determines that a portion of the detected line segment contained in the input image sub-block has a line segment defect, the processing device reports the type of the line segment defect and the location of the image sub-block containing the line segment defect.

[0007] In one embodiment, the above-mentioned equipment for detecting panel line segment defects also includes a lighting device and a plurality of closed partitions. The lighting device and the image capture device are arranged on opposite sides of the panel to be tested and project light onto the panel to be tested. These closed partitions are combined to form an opaque area surrounding the image capture device, the panel to be tested and the lighting device.

[0008] In one embodiment, the processing device further finds at least one existing line segment from each of a plurality of ideal sample panel images, divides the existing line segments into a plurality of position groups according to their positions, and stores the average position of at least one existing line segment in each position group into the sample storage device as part of the sample line segment position.

[0009] From another perspective, the description of the present invention also provides a method for detecting line segment defects of a panel, which is suitable for confirming whether there are line segment defects on the panel to be tested. This method first photographs the panel to be tested to obtain an initial image, and captures the image of the panel to be tested corresponding to the panel to be tested from the initial image; then, by first obtaining the sample data stored in advance, according to the sample line segment position included in the sample data, the image corresponding to the sample line segment position on the image of the panel to be tested is set as the image of the line segment to be tested; finally, the image of the line segment to be tested is checked using a neural network classification program to determine whether the line segment to be tested has a line segment defect.

[0010] In one embodiment, the above-mentioned step of using a neural network classification program to check the image of the line segment to be tested to determine whether the line segment to be tested has a line segment defect, includes: dividing the image block containing the line segment to be tested in the image of the panel to be tested into a plurality of image sub-blocks; and using each image sub-block as input data of the neural network classification program so that the neural network classification program determines whether the portion of the line segment to be tested contained in the input image sub-block has a line segment defect.

[0011] In one embodiment, when the neural network program determines that a portion of the detected line segment contained in the input image sub-block has a line segment defect, it further reports the type of the line segment defect and the position of the image sub-block containing the line segment defect.

[0012] In one embodiment, the following steps are performed before photographing the panel to be tested to obtain the initial image: finding at least one existing line segment from each of a plurality of ideal sample panel images; dividing the existing line segments into a plurality of position groups according to the positions of the found existing line segments, wherein each position group includes at least one existing line segment; averaging the positions of the existing line segments in each position group to obtain a corresponding average result; and storing each average result as a part of the sample line segment position.

[0013] In one embodiment, the above-mentioned method of finding at least one existing line segment from each of a plurality of ideal sample panel images includes: performing the following steps on each ideal sample panel image, wherein each ideal sample panel image includes a plurality of pixels, and a plane where each ideal sample panel image is located is formed by extending a first axis and a second axis: converting the image value of each pixel in the ideal sample panel image into a first value or a second value; in this ideal sample panel image, for each coordinate value of the first axis, calculating a first calculation number of the image value of the pixel having the first axis coordinate value as a first value, and when the first calculation number exceeds a first preset value, setting the existence of a first line segment to be integrated at the first axis coordinate value, and making the first line segment to be integrated a part of the above-mentioned existing line segment; and, in this ideal sample panel image, for each coordinate value of the second axis, calculating a second calculation number of the image value of the pixel having the second axis coordinate value as a first value, and when the second calculation number exceeds a second preset value, setting the existence of a second line segment to be integrated at the second axis coordinate value, and making the second line segment to be integrated a part of the above-mentioned existing line segment.

[0014] According to the above, the device and method for detecting panel line segment defects provided by the description of the present invention can automatically locate the position of the panel line segment to be detected on the panel through the image of the panel, and can further send the image of the located panel line segment to the neural network classification program so that the neural network classification program trained in advance can determine whether there is a defect in the panel line segment. Therefore, the technology provided by the present invention can not only reduce the manpower required for determining defects, but also reduce the problem of uneven panel quality caused by different human inspection standards. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 Schematic diagram of the appearance of a device for detecting line segment defects of a panel according to an embodiment of the present invention.

[0016] Figure 2 4 is a circuit block diagram of a device for detecting panel line segment defects according to an embodiment of the present invention.

[0017] Figure 3 The figure is a flow chart of a method for detecting line segment defects of a panel according to an embodiment of the present invention.

[0018] Figure 4A FIG. 4 is a schematic diagram of an ideal sample panel image used according to an embodiment of the present invention.

[0019] Figure 4B for Figure 4A A bar graph showing the relationship between the X-axis coordinate and the number of black pixels.

[0020] Figure 4C for Figure 4A A bar graph showing the relationship between the Y-axis coordinate and the number of black pixels.

[0021] Figure 5 FIG. 4 is a flow chart of establishing sample data according to an embodiment of the present invention.

[0022] Figure 6 FIG. 4 is a schematic diagram of dividing an image block into a plurality of image sub-blocks according to an embodiment of the present invention.

[0023] Figure 7 Schematic diagram of the appearance of a device for detecting line segment defects of a panel according to an embodiment of the present invention.

[0024] Among them, the simple description of the symbols in the accompanying drawings is as follows:

[0025] 10, 70: equipment for detecting line defects of panel; 40: image of ideal sample panel; 60: image of panel to be tested; 100: image acquisition device; 110: processing device; 115: neural network classification program; 120: sample storage device; 125: sample data; 130: panel to be tested; 140: lighting device; 155: shooting area; 400, 402, 404, 406, 410, 412, 414: existing line defects Segments; 600, 602, 604, 606, 610, 612, 614: line segments to be measured; 602-1, 602-2, 602-N, 612-1, 612-2, 612-M: image sub-blocks; 700: closed partition; n: number; S300~S308: implementation steps of an embodiment of the present invention; S500~S510: implementation steps of an embodiment of the present invention when establishing sample data; X, Y: axes. DETAILED DESCRIPTION

[0026] Please also refer to Figure 1 and Figure 2 ,in, Figure 1 is a schematic diagram of the appearance of a device for detecting line segment defects of a panel according to an embodiment of the present invention, Figure 2 for Figure 1In the present embodiment, the device 10 comprises an image capturing device 100, a processing device 110 and a sample storage device 120, wherein the processing device 110 is electrically connected to the image capturing device 100 and the sample storage device 120 respectively so that the processing device 110 can perform data transmission operations with the image capturing device 100 and the sample storage device 120 respectively.

[0027] In order to enable those skilled in the art to more clearly understand the operation of the device 10, the following will be combined with Figure 1 , Figure 2 and Figure 3 To explain in detail, Figure 3 The figure is a flow chart of a method for detecting line segment defects of a panel according to an embodiment of the present invention.

[0028] like Figure 1 As shown, the shooting area 155 is the range that the image capture device 100 can cover when shooting, and the panel to be tested 130 is placed in the shooting area 155. Before actually detecting the panel line segment defects, a preset control program can be used to control the processing device 110 or other suitable control devices to operate the image capture device 100 to shoot the shooting area 155, thereby obtaining an image of the entire shooting area 155 (hereinafter referred to as the initial image) (step S300). It should be noted that although the size of the shooting area 155 in this embodiment is larger than the size of the panel to be tested 130, it is actually still possible to make the size of the shooting area 155 smaller than the size of the panel to be tested 130 and obtain the entire image of the panel to be tested 130 by moving the position of the panel to be tested 130 or the image capture device 100 and performing multiple shootings.

[0029] The initial image captured by the image capture device 100 is transmitted to the processing device 110 for further processing. After obtaining the initial image from the image capture device 100, the processing device 110 then extracts the portion corresponding to the panel to be tested 130 from the initial image to obtain the panel to be tested image (step S302). Figure 1Taking the embodiment shown as an example, since the size of the shooting area 155 is larger than the size of the panel to be tested 130, there will be images that are not the panel to be tested 130 in the initial image. In this case, the processing device 110 can remove the image portion that is not the panel to be tested 130 or extract the image portion related to the panel to be tested 130 by executing a related image processing program (such as a neural network program U-Net or other programs) to finally generate an image of the panel to be tested. In another embodiment, if the size of the shooting area 155 is smaller than the size of the panel to be tested 130, the processing device 110 can also combine multiple initial images together by executing a related image processing program to generate an image of the panel to be tested required for subsequent processing.

[0030] In addition to the above-mentioned pre-processing procedures, the processing device 110 also obtains the sample data 125 stored in the sample storage device 120 (step S304). It can be seen that the sample data 125 should be established before the actual detection of line segment defects. In order to establish the sample data 125, the processing device 110 must first collect information related to the position of the line segment to be detected. Please refer to Figure 4A and Figure 5 ,in, Figure 4A is a schematic diagram of an ideal sample panel image used according to an embodiment of the present invention, Figure 5 FIG. 1 is a flow chart of establishing sample data according to an embodiment of the present invention. By controlling the image capture device 100 to capture an ideal sample panel once and performing the aforementioned steps S300 and S302, an ideal sample panel image corresponding to the ideal sample panel can be obtained (step S500). In this embodiment, Figure 4A The ideal sample panel image 40 shown includes a plurality of pixels, and the plane in which it is located is extended by the X-axis (hereinafter also referred to as the first axis) and the Y-axis (hereinafter also referred to as the second axis). Through the image processing technology, the processing device 110 can find the existing line segments 400, 402, 404 and 406 extending along the Y-axis and the existing line segments 410, 412 and 414 extending along the X-axis from the ideal sample panel image 40 (step S502). The position information of the existing line segments 400-414 found in step S502 can be directly used for subsequent processing, or, as Figure 5As shown, after performing image processing on an ideal sample panel image and finding the position information of the corresponding existing line segments, it can be further confirmed whether it is necessary to continue to obtain other more ideal sample panel images (step S504). If it is necessary to obtain other ideal sample panel images, the processing device 110 will repeat steps S500 to S502 to obtain another set of position information of existing line segments. Conversely, if it is not necessary to obtain other ideal sample panel images, the processing device 110 will enter step S506 with the existing position information of all existing line segments and continue the subsequent processing.

[0031] It is worth mentioning that, although the existing line segments can be found by manual marking in step S502, the present invention proposes to improve the process of finding the existing line segments by automating it. In order to find the existing line segments without manual work, in one embodiment, the processing device 110 can first binarize the image value of each pixel in the ideal sample panel image, that is, adjust the ideal sample panel image that originally appears in grayscale or color when taken to an image with only black and white colors. Next, the processing device 110 begins to calculate the number of pixels that appear on the same X-axis coordinate and represent the color of the existing line segments (e.g., black) (hereinafter referred to as the first calculated number), and calculates the number of pixels that appear on the same Y-axis coordinate and represent the color of the existing line segments (e.g., black) (hereinafter referred to as the second calculated number). For example, please refer to Figure 4B , where the horizontal axis represents the coordinate value of the X axis, and the vertical axis represents the number of black pixels n, so Figure 4B The heights of the four straight lines from left to right in the figure respectively represent the number of black pixels of the existing line segment 400, the number of black pixels of the existing line segment 402, the number of black pixels of the existing line segment 404, and the number of black pixels of the existing line segment 406; please also refer to Figure 4C , where the horizontal axis represents the coordinate value of the Y axis, and the vertical axis also represents the number of black pixels n, so Figure 4C The heights of the three straight lines from left to right in 4 represent the number of black pixels of the existing line segment 414 , the number of black pixels of the existing line segment 412 , and the number of black pixels of the existing line segment 410 , respectively.

[0032] exist Figure 4A In the embodiment shown, for each X-axis coordinate except for the position where the existing line segments 400-406 are located, black pixels that exist due to the existing line segments 410-414 may be calculated, and black pixels that are generated out of thin air due to photographic distortion or image processing errors may also be calculated. Figure 4BAfter the first calculated number corresponding to each X-axis coordinate, the parts of the existing line segments that are not extended along the Y-axis must be excluded. Therefore, a preset value (hereinafter referred to as the first preset value) can be set in advance and only the coordinates where the first calculated number exceeds the first preset value are set as the positions where the existing line segments exist. Similarly, for each Y-axis coordinate except the positions where the existing line segments 410 to 414 are located, black pixels that exist due to the existing line segments 400 to 406 may be calculated, and black pixels that are generated out of thin air due to photographic distortion or image processing errors may also be calculated. Therefore, when obtaining Figure 4C After the second calculated number corresponding to each Y-axis coordinate, the parts of the existing line segments that are not extended along the X-axis must be excluded. Therefore, a preset value (hereinafter referred to as the second preset value) can be set in advance and only the coordinates where the second calculated number exceeds the second preset value are set as the positions where the existing line segments exist.

[0033] Through the above-mentioned approach, the present invention can automate the process of finding existing line segments and reduce the manpower required.

[0034] Please refer back Figure 5. After the operations of steps S500 to S504, the processing device 110 first groups the existing line segments obtained from the same ideal sample panel image in step S506. When grouping, the processing device 110 will classify different existing line segments into different position groups. For example, as mentioned above, existing line segments 400 to 414 can be obtained from the ideal sample panel image 40, so the processing device 110 will classify different existing line segments 400 to 414 into different position groups, for example, the existing line segment 400 is classified into the first position group, the existing line segment 402 is classified into the second position group, the existing line segment 404 is classified into the third position group, the existing line segment 406 is classified into the fourth position group, the existing line segment 410 is classified into the fifth position group, the existing line segment 412 is classified into the sixth position group, and the existing line segment 414 is classified into the seventh position group. In addition to the positions of the existing line segments, the positions included in each position group may also include positions within a preset range of distance from each existing line segment. For example, the first position group may include an interval extending from the leftmost end of the ideal sample panel image 40 to the right to the midline between the existing line segment 400 and the existing line segment 402; the second position group may include an interval extending from the midline between the existing line segment 400 and the existing line segment 402 to the right to the midline between the existing line segment 402 and the existing line segment 404; the third position group may include an interval extending from the midline between the existing line segment 402 and the existing line segment 404 to the right to the midline between the existing line segment 404 and the existing line segment 406; and the fourth position group may include an interval extending from the midline between the existing line segment 404 and the existing line segment 406 to the right to the rightmost end of the ideal sample panel image 40. Similarly, the fifth position group may include an interval extending from the uppermost end of the ideal sample panel image 40 to the center line between the existing line segment 410 and the existing line segment 412; the sixth position group may include an interval extending from the center line between the existing line segment 410 and the existing line segment 412 to the center line between the existing line segment 412 and the existing line segment 414; the seventh position group may include an interval extending from the center line between the existing line segment 412 and the existing line segment 414 to the lowermost end of the ideal sample panel image 40.

[0035] It should be noted that the size of the interval covered by each position group may be defined in other ways, such as including an interval of 10 pixels near the position of each existing line segment, etc. Those skilled in the art may make appropriate adjustments as needed.

[0036] Next, when there is only one ideal sample panel image that needs to be operated in step S506, the processing device 110 can bring the relevant information of each position group obtained at this time (the position of the existing line segment in each position group) into step S508 for subsequent processing. Conversely, when there are many ideal sample panel images that need to be operated in step S506, the processing device 110 must perform the above classification operation once for each different ideal sample panel image. As known to those skilled in the art, since different ideal sample panel images obtained by shooting multiple ideal sample panels are to be processed, it can be expected that the corresponding existing line segments in these ideal sample panel images should fall at similar positions. For example, the existing line segment 400 found in the first ideal sample panel image and the existing line segment 400 found in the second ideal sample panel image theoretically appear at the same position; even considering the error phenomenon that may occur during the image processing process, the existing line segment 400 found in the first ideal sample panel image should not be too far away from the existing line segment 400 found in the second ideal sample panel image. Therefore, after performing the aforementioned position group classification on each ideal sample panel image, the processing device 110 may classify mutually close position groups obtained from different ideal sample panel images into the same position group (Position Group Set).

[0037] In different embodiments, the processing device 110 may adopt any appropriate classification criteria to classify different position groups into the same position group, for example, the coverage of the first position group of the first ideal sample panel image is compared with the coverage of the first to fourth position groups in the second ideal sample panel image that also include existing line segments extending along the Y axis, and the position group in the second ideal sample panel image that overlaps the most with the coverage of the first position group of the first ideal sample panel image is classified into the same position group as the first position group of the first ideal sample panel image. Similar operations are also applied to each position group of each ideal sample panel image in step S506 to thereby classify the existing line segments into multiple position groups.

[0038] After the processing of step S506, each position group will contain at least one existing line segment. Next, the processing device 110 will average the positions of the existing line segments in each position group in step S508 to obtain the corresponding average result (if there is only one existing line segment, the average result is naturally the position of the existing line segment), and store the obtained average result in the sample data 125 so that each average result becomes a sample line segment position or a part of the sample data 125 (step S510). Obviously, when there is only one average result (that is, there is only one position group), the sample line segment position (or sample data 125) will have only one content; and when there are N average results, the sample line segment position (or sample data 125) will have N content.

[0039] Please come back Figure 3 , after obtaining Figure 5 After obtaining the sample data 125 through the operation process, the processing device 110 sets the position specified by each content corresponding to the sample line segment position in the image of the panel to be tested and the image block within a specific distance therefrom as a corresponding image of the line segment to be tested according to the sample data 125 (step S306). After obtaining the image of each line segment to be tested, the processing device 110 can input the image of each line segment to be tested into the neural network classification program 115 (for example, but not limited to, the neural network program MobileNetV2) to determine whether there is a specific line segment defect in the image of each line segment to be tested through the neural network classification program 115 (step S308).

[0040] The neural network classification program 115 should be trained before judging line defects. Once the neural network classification program 115 can complete the judgment training for certain types of line defects (e.g., line width variation, voids, broken lines, etc.), after the image of the line segment to be tested is input into the neural network classification program 115, the neural network classification program 115 can judge whether the line segment to be tested has these specific types of line defects according to fixed standards.

[0041] It should be noted that the aforementioned image of the line segment to be tested is not limited to being completely input into the neural network classification program 115. In different embodiments, the processing device 110 may also divide the image block containing the line segment to be tested into a plurality of image sub-blocks, and use each image sub-block as input data of the neural network classification program 115 so that the neural network classification program 115 can analyze this image sub-block, thereby determining whether the part of the line segment to be tested contained in this image sub-block has a specific line segment defect. Figure 6 A schematic diagram of dividing an image block including a line segment to be detected into a plurality of image sub-blocks according to an embodiment of the present invention. Figure 6 The panel image 60 to be tested includes a plurality of line segments 600, 602, 604, 606, 610, 612 and 614 to be tested. The processing device 110 may divide the image block including the line segments to be tested into a plurality of image sub-blocks. Here, only the line segments 602 and 612 to be tested are used as examples for description. Figure 6 As shown, the processing device 110 can divide the image block containing the line segment 602 to be tested into a plurality of image sub-blocks 602-1, 602-2, ..., 602-N, and divide the image block containing the line segment 612 to be tested into a plurality of image sub-blocks 612-1, 612-2, ..., 612-M (blocks shown by dotted lines in the figure). When the processing device 110 uses the neural network classification program 115 to determine that a portion of the line segment to be tested contained in a certain image sub-block has a specific line segment defect, the type of the line segment defect and the position of the image sub-block containing the line segment defect can be reported for subsequent processing. For example, if the processing device 110 reports that there is a line segment defect with a broken line at the position of the image sub-block 612-6 (not shown), the line segment can be laser repaired at the position of the image sub-block 612-6 in the panel to be tested to improve the panel yield. Those skilled in the art will appreciate that, in embodiments in which image sub-blocks are used as input data for the neural network classification program 115, when training the neural network classification program 115, it is necessary to train it to accept image sub-blocks (rather than the image of the entire line segment to be measured) as input data, and each image sub-block used as input data for the neural network classification program 115 must also be scaled to the same image size, which will not be elaborated herein.

[0042] It can be seen from the above description that the present invention has the ability to automatically determine line segment defects.

[0043] Please refer to Figure 1 The device 10 may further include an illumination device 140. In this embodiment, the illumination device 140 and the image capturing device 100 are respectively disposed on opposite sides of the panel to be tested 130 and project light onto the panel to be tested 130. Further, in order to reduce the image problem caused by external light, Figure 7 In the embodiment shown, a light-proof closed partition 700 is used to form a Figure 1 The image capture device 100, the panel to be tested 130 and the light-proof area surrounded by the lighting device 140 are shown. In this way, since the light used by the image capture device 100 when shooting is only the light provided by the lighting device 140, when the device 70 is used to determine line segment defects, the problem derived from the interference of external light and shadow changes on the image can be greatly reduced.

[0044] In summary, the device and method for detecting panel line segment defects provided in the above content can automatically locate the position of the panel line segment to be detected on the panel through the image of the panel, and can further send the image of the located panel line segment to the neural network classification program so that the neural network classification program trained in advance can determine whether there is a defect in the panel line segment. Therefore, the technology provided by the present invention can not only reduce the manpower required for determining defects, but also reduce the problem of uneven panel quality caused by different human inspection standards.

[0045] The above description is only a preferred embodiment of the present invention, but it is not intended to limit the scope of the present invention. Any person familiar with this technology can make further improvements and changes on this basis without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be based on the scope defined by the claims of this application.

Claims

1. A device for detecting line segment defects of a panel, suitable for confirming whether there are line segment defects on the panel to be tested, the device is characterized by comprising: An image capture device is used to capture an area where the panel to be tested is placed to obtain an initial image; A sample storage device stores sample data, the sample data including sample line segment positions; and A processing device electrically connected to the image capturing device and the sample storage device, The processing device receives the initial image from the image capture device and obtains the image of the panel to be tested corresponding to the panel to be tested from the initial image, and the processing device also obtains the sample data from the sample storage device to set the image at the position corresponding to the sample line segment on the panel to be tested as the image of the line segment to be tested. The processing device executes a neural network classification program to check the image of the line segment to be tested and determine whether the line segment to be tested has the line segment defect.

2. The device as described in claim 1, wherein the processing device divides the image block containing the line segment to be tested in the image of the panel to be tested into a plurality of image sub-blocks, and uses each of the plurality of image sub-blocks as input data for the neural network classification program so that the neural network classification program determines whether the portion of the line segment to be tested contained in the input image sub-block has the line segment defect.

3. The apparatus as claimed in claim 2, wherein when the neural network classification program determines that the portion of the line segment to be tested included in the input image sub-block has the line segment defect, the processing device reports the type of the line segment defect and the position of the image sub-block including the line segment defect.

4. The device as described in claim 1 further includes a lighting device and a plurality of closed partitions, wherein the lighting device and the image capture device are arranged on opposite sides of the panel to be tested and project light onto the panel to be tested, and the plurality of closed partitions are combined to form an opaque area surrounding the image capture device, the panel to be tested and the lighting device.

5. The apparatus as described in claim 1, wherein the processing device further finds at least one existing line segment from each of the plurality of ideal sample panel images, divides the at least one existing line segment into a plurality of position groups according to the position of the at least one existing line segment, and stores the average result of the position of the at least one existing line segment in each of the plurality of position groups into the sample storage device as a part of the sample line segment position.

6. A method for detecting line segment defects on a panel, suitable for confirming whether there are line segment defects on the panel to be tested, the method is characterized by comprising: photographing the panel to be tested to obtain an initial image; Capturing a panel image corresponding to the panel to be tested from the initial image; Obtaining previously stored sample data, wherein the sample data includes a sample line segment position; According to the position of the sample line segment, setting the image corresponding to the position of the sample line segment on the image of the panel to be tested as the image of the line segment to be tested; and A neural network classification program is used to check the image of the line segment to be tested to determine whether the line segment to be tested has the line segment defect.

7. The method of claim 6, wherein using the neural network classification program to check the image of the line segment to be tested to determine whether the line segment to be tested has the line segment defect comprises: Dividing an image block containing the line segment to be tested in the image of the panel to be tested into a plurality of image sub-blocks; as well as Each of the plurality of image sub-blocks is used as input data of the neural network classification program so that the neural network classification program determines whether the portion of the line segment to be detected included in the input image sub-block has the line segment defect.

8. The method of claim 7, wherein when the neural network classification program determines that the portion of the line segment to be detected included in the input image sub-block has the line segment defect, the method further comprises: The type of the line defect and the position of the image sub-block containing the line defect are reported.

9. The method according to claim 6, wherein before photographing the panel to be tested to obtain the initial image, the method further comprises: Find at least one existing line segment from each of the plurality of ideal sample panel images; Dividing the at least one existing line segment into a plurality of position groups according to the position of the at least one existing line segment, wherein each of the plurality of position groups includes the at least one existing line segment; Averaging the position of the at least one existing line segment in each of the plurality of position groups to obtain a corresponding average result; as well as Each of the average results is stored as a portion of the sample line segment position.

10. The method of claim 9, wherein finding the at least one existing line segment from each of the plurality of ideal sample panel images comprises: Perform the following steps on each of the plurality of ideal sample panel images, wherein each of the plurality of ideal sample panel images includes a plurality of pixels, and a plane where each of the plurality of ideal sample panel images is located is extended from a first axial direction and a second axial direction: Converting the image value of each of the plurality of pixels in the ideal sample panel image into a first value or a second value; In the ideal sample panel image, for each coordinate value of the first axis, the image values ​​of the plurality of pixels having the first axis coordinate value are calculated as a first calculation number of the first value, and when the first calculation number exceeds a first preset value, a first line segment to be integrated is set to exist at the first axis coordinate value, and the first line segment to be integrated is made a part of the at least one existing line segment; as well as In the ideal sample panel image, for each coordinate value of the second axis, the image values ​​of the multiple pixels that also have the second axis coordinate value are calculated as a second calculated number of the first value. When the second calculated number exceeds a second preset value, it is set that a second line segment to be integrated exists at the second axis coordinate value, and the second line segment to be integrated is made a part of the at least one existing line segment.

Images