Panel detection method and device based on multi-depth height detection shooting

Abstract

The invention provides a panel detection method based on multi-depth height detection shooting. The method comprises the steps that first height information of a sampling point in a detection view field range on the surface plane of a panel is acquired through a height measurement probe; clustering the sampling points according to the corresponding first height information to obtain sub-regions related to height distribution in the detection view field range, wherein the information of one sub-region has second height information; sequentially moving the image detection sensor and / or adjusting the shooting focal length of the image detection sensor along the optical axis direction of the image detection sensor according to the second height information corresponding to the sub-regions, and shooting the detection region to obtain one or more component images corresponding to the sub-regions; and respectively taking the area images of the corresponding sub-areas in the component images, and combining the area images into a detection area image for detection. According to the panel detection method, the accuracy of AOI can be improved.

Description

Technical Field

[0001] The present invention relates to the field of detection, and particularly to a panel detection method and device based on multi-depth height detection shooting. Background Art

[0002] In the field of AOI (Automated Optical Inspection) detection in the panel Module section, with the update and replacement of process technologies, higher precision requirements are put forward for the detection of special types of defects. For example, the original defect detection requirement at the micron level has now been upgraded to the sub-micron level, which makes the magnification of the used lens higher and higher, and correspondingly, the depth of field range of the lens is gradually compressed.

[0003] However, it is difficult to ensure a good flatness on the panel surface. Therefore, in order to ensure the clarity of the photo, it is necessary to measure the height of the panel surface, that is, to detect the height of the panel surface. The result of the height detection will guide the detection probe (image detection sensor) to adjust the Z-axis height in a timely manner during the detection process, so that the probe can focus on the area within the detection field of view on the panel surface.

[0004] In the traditional technology, only the height information of the center point position within the detection field of view of the probe is used to adjust the Z-axis height or the probe. This causes the situation that when the height within the detection field of view is uneven, the image captured by the probe will be locally unclear, which affects the accuracy of the panel detection. Summary of the Invention

[0005] To solve the above problems, the present invention provides a panel detection method based on multi-depth height detection shooting, including:

[0006] Obtaining first height information of sampling points within a detection field of view on the panel surface by a height measurement probe, where the number of sampling points included within the detection field of view is multiple;

[0007] Clustering the sampling points according to the corresponding first height information to obtain sub-regions related to the height distribution within the detection field of view, and one sub-region information has second height information;

[0008] Taking a photo of the detection field of view by an image detection sensor for detection:

[0009] Moving the image detection sensor along the optical axis direction of the image detection sensor and / or adjusting the shooting focal length of the image detection sensor in sequence according to the second height information corresponding to each sub-region, taking a photo of the detection area, and obtaining one or more component images corresponding to the sub-regions;

[0010] In the component images, region images of corresponding sub-regions are respectively taken and merged into a detection region image for detection.

[0011] In one embodiment, the height measurement probe is a linear height measurement probe.

[0012] In one embodiment, the step of sequentially moving the image detection sensor and / or adjusting the shooting focal length of the image detection sensor along the optical axis direction of the image detection sensor according to the second height information corresponding to each sub-region includes:

[0013] Excluding the sub-regions where the second height information exceeds the first threshold range.

[0014] In one embodiment, the step of excluding the sub-regions where the second height information exceeds the first threshold range includes:

[0015] If the size of the sub-region where the second height information exceeds the first threshold range is less than the third threshold, determining that the sub-region is a defective region.

[0016] In one embodiment, the step of excluding the sub-regions where the second height information exceeds the first threshold range includes:

[0017] If the size of the sub-region where the second height information exceeds the first threshold range is greater than the third threshold, determining that the sub-region is a feature structure region.

[0018] In one embodiment, the step of clustering the sampling points according to the corresponding first height information to obtain sub-regions related to height distribution within the detection field of view includes:

[0019] Setting a height interval for clustering according to the imaging range of the image detection sensor.

[0020] To solve the above problems, the present invention also provides a panel detection device based on multi-depth height detection shooting, including:

[0021] A height detection module, configured to obtain first height information of sampling points within a detection field of view on the surface plane of the panel through a height measurement probe, and the number of sampling points included in the detection field of view is multiple;

[0022] A sampling point clustering module, configured to cluster the sampling points according to the corresponding first height information to obtain sub-regions related to height distribution within the detection field of view, and one sub-region information has second height information;

[0023] The component image acquisition module is configured to move the image detection sensor along the optical axis direction of the image detection sensor and / or adjust the shooting focal length of the image detection sensor in sequence according to the second height information corresponding to each of the sub-regions, take pictures of the detection region, and obtain one or more component images corresponding to the sub-regions;

[0024] The image merging and detection module is configured to merge the region images of the corresponding sub-regions in the component images into a detection region image for detection.

[0025] In one embodiment, the height measurement probe is a linear height measurement probe.

[0026] In one embodiment, the component image acquisition module is further configured to exclude the sub-regions whose second height information exceeds the first threshold range.

[0027] In one embodiment, the component image acquisition module is further configured to determine that the sub-region is a defective region when the size of the sub-region whose second height information exceeds the first threshold range is smaller than the second threshold.

[0028] In one embodiment, the component image acquisition module is further configured to determine that the sub-region is a feature structure region when the size of the sub-region whose second height information exceeds the first threshold range is larger than the third threshold.

[0029] In one embodiment, the sampling point clustering module is configured to set a clustering height interval according to the imaging range of the image detection sensor.

[0030] Compared with the prior art, the panel detection method and device based on multi-depth height detection of the present application have the following beneficial effects.

[0031] The panel detection method based on multi-depth height detection of the present invention performs height sampling of multiple sampling points within the detection field of view of an image detection sensor (i.e., a detection probe), and clusters them into sub-regions of multiple heights within the detection field of view. The image detection sensor can take component images of multiple sub-regions only when moving or focusing in the direction perpendicular to the panel surface, and then merge the component images into an image of the entire detection field of view. Since each component image is taken within the imaging range of the image detection sensor, the merged image of the entire detection field of view is clear in each region, thus ensuring the clarity of the image collected by the image detection sensor and improving the detection accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a schematic diagram of an AOI detection system in an embodiment of the present application;

[0033] Figure 2 It is the working principle diagram of the optical alignment process of the AOI detection system;

[0034] Figure 3 It is the schematic diagram of the height detection process in the traditional technology;

[0035] Figure 4 It is the schematic diagram of the position of the singular point in the height detection process in the traditional technology;

[0036] Figure 5 It is the flowchart of a panel detection method based on multi-depth height detection shooting in an embodiment of the present application;

[0037] Figure 6 It is the schematic diagram of the height detection process for multiple sampling points within the detection field of view in an embodiment of the present application;

[0038] Figure 7 It is the schematic diagram of determining the surface defects of the panel specimen through height detection in an embodiment of the present application;

[0039] Figure 8 It is the schematic diagram of determining the surface defects of the panel specimen through height detection in an embodiment of the present application;

[0040] Figure 9 It is the schematic diagram of the AOI detection system using a linear height measurement probe in an embodiment of the present application;

[0041] Figure 10 It is the schematic diagram of a panel detection device based on multi-depth height detection shooting in an embodiment of the present application;

[0042] Figure 11 It is the schematic diagram of 4 typical forms of height differences on the panel surface in an embodiment of the present application;

[0043] Figure 12 It is the structural block diagram of a panel detection device based on multi-depth height detection shooting in an embodiment of the present application. Detailed implementation manners

[0044] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0045] It should be noted that if there are directional indications (such as up, down, left, right, front, back...) involved in the embodiments of this application, then such directional indications are only used to explain the relative positional relationship, movement conditions, etc. between components in a specific posture (as shown in the drawings). If this specific posture changes, then the directional indications will also change accordingly.

[0046] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of this application, then such descriptions of "first", "second", etc. are only for descriptive purposes and should not be construed as indicating or implying their relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In addition, the technical solutions between various embodiments can be combined with each other, but it must be based on the ability of those of ordinary skill in the art to implement. When the combination of technical solutions results in contradictions or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0047] As Figure 1 shown, Figure 1 a kind of AOI detection system for a panel is shown. In Figure 1 , the panel specimen 10 is placed flat on the guide rail 60 and translated, and sequentially passes through three processes of optical alignment, height detection, and detection from left to right. Among them, optical alignment depends on the alignment probe 20, which respectively takes pictures of the Mark points at the two corners of the short side of the product. As Figure 2 shown, when the panel specimen is in an offset state, the position of the Mark captured by the alignment probe will shift accordingly. By comparing with the position in the standard state, the offset amount of the panel specimen can be obtained according to the coordinate system conversion relationship, and then the positioning parameters can be corrected.

[0048] Figure 1 The second step in it is the height detection process. The height detection probe 30 in the traditional height detection process uses a point height detection probe, that is, it only detects the height value of one point at a time. And the selection of the detection position should be determined according to the detection field of view range of the detection probe in the third step. This is because the detection accuracy is relatively high, reaching the sub-micron level, so that the detection field of view range of each detection probe only covers a small area on the surface of the panel specimen, and multiple detection probes 40 need to be set or the detection probe 40 needs to be moved to superimpose the detection field of view range to cover the entire surface of the panel specimen 10. In Figure 1 , the alignment probe 20, the height detection probe 30, and the detection probe 40 are respectively arranged on the probe slide rails 52, 54, and 56.

[0049] In the traditional technology, when the detection probe 40 is translated according to the path planning to sequentially photograph the specified detection areas on the surface of the panel specimen, the center point of each specified detection area or a certain specific point therein is used as the sampling point, that is, the detection position of the point height detection probe. By adjusting the height or focal length of the detection probe in this way, it can be ensured that each detection area is in clear focus.

[0050] However, the data stability of the point height detection probe in the traditional technology is not high and is easily affected by singular points (for example, there are defects just at the center point, resulting in too low or too high height), which may cause serious defocusing. In addition, even if the focusing at the center position is relatively ideal, the height information of other positions within the detection area cannot be obtained. Therefore, the situation where the center of the detection area is clear while the periphery is blurred may also occur, which will cause the detection probe to be unable to capture a comprehensive and clear image, thus affecting the detection accuracy of AOI.

[0051] As Figure 3 shown, for a certain area on the panel specimen, the detection field of view range of the detection probe 40 is small and cannot cover the entire area. Then, the area is pre-divided into 9 detection areas, and the center point of each detection area is the sampling point of the height detection probe 30. However, it can be seen that if the detection area contains both a normal height surface and a sunken surface, when the sampling point is located on the normal height surface, the sunken surface may exceed the focal length range of the detection probe and cause blurring; when the sampling point is located on the sunken surface, the normal height surface may exceed the focal length range of the detection probe and cause blurring. Again, as Figure 4 shown, for a certain detection area on the panel specimen, if there is an abnormal protrusion and the sampling point just happens to be at the top of the abnormal protrusion, it will cause the detection probe 40 to be severely defocused. In these two cases, the detection probe cannot capture a clear image of the detection area, and the accuracy of AOI will be affected.

[0052] To solve the problem that the height detection process in the traditional technology may affect the detection accuracy of AOI, an embodiment of the present invention proposes a panel detection method based on multi-depth height detection, as Figure 5 shown, including:

[0053] Step S102: Obtain the first height information of the sampling points within a detection field of view range on the flat surface of the panel through a height measurement probe (i.e., the height detection probe), and the number of sampling points included in this detection field of view range is multiple. The detection field of view range is the range covered by the field of view when the image detection sensor takes a picture, and the size of the image taken by the image detection sensor is the size of the detection field of view range.

[0054] Refer to Figure 1 and Figure 6As shown, in this embodiment, the panel specimen is horizontally placed on the guide rail and translated from left to right. When the panel specimen enters the height detection procedure at the position of the height detection probe, the height detection probe collects the first height information at multiple sampling points on the upper surface of the panel specimen, and a plurality of sampling points are included within the size range of one detection field of view. That is to say, if the size of the detection field of view of the image detection sensor is M×N, the distance between two adjacent sampling points is less than M and less than N, so that a plurality of sampling points are included within the range of M×N. As Figure 6 In the figure, 9 sampling points are included within one detection field of view. That is to say, the height detection probe will detect the first height information 9 times at 9 positions within one detection field of view.

[0055] Step S104: Cluster the sampling points according to the corresponding first height information to obtain sub-regions related to the height distribution within the detection field of view, and one sub-region information has the second height information.

[0056] In Figure 6 's example, the first height information at the positions of 6 sampling points out of 9 sampling points is D1, and the first height information at the positions of 3 sampling points is D2. According to the distribution density of the sampling points, the area covered by each sampling point is set, and then the sub-region with D1 as the height and the sub-region with D2 as the height can be obtained, that is Figure 6 In the figure, the sub-region composed of 1, 2, 3, 4, 7 (abbreviated as the 12347 region) has the corresponding second height information of D1; the sub-region composed of 5, 6, 8, 9 (abbreviated as the 5689 region) has the corresponding second height information of D2.

[0057] Step S106: Move the image detection sensor along the optical axis direction of the image detection sensor and / or adjust the shooting focal length of the image detection sensor in sequence according to the second height information corresponding to each sub-region, take pictures of the detection region, and obtain one or more component images corresponding to the sub-region.

[0058] The image detection sensor is arranged on the slide rail and can move in the XY plane parallel to the surface of the panel specimen to switch the detection field of view, or can move up and down in the Z direction perpendicular to the surface of the panel specimen for the same detection field of view to adjust the distance between its lens and the surface of the panel specimen.

[0059] In the above example, after the image detection sensor moves on the XY plane to cover the detection field of view, the image detection sensor can be first moved along the optical axis direction of the image detection sensor and / or the shooting focal length of the image detection sensor can be adjusted according to the first height information D1, so that the area 12347 is clearly visible in the image, and then the first component image corresponding to the area 12347 is captured; then, according to the second height information D2, the image detection sensor is moved along the optical axis direction of the image detection sensor and / or the shooting focal length of the image detection sensor is adjusted, so that the area 5689 is clearly visible in the image, and then the second component image corresponding to the area 5689 is captured. During this period, the image detection sensor does not need to move on the XY plane.

[0060] It should be noted that the sub-regions obtained by clustering can be discrete rather than continuous, that is, the sub-regions corresponding to the same second height information can be discretely distributed within the detection field of view.

[0061] Reference Figure 7 shown in Figure 7 contains 5 sub-regions, and the corresponding second height information is Z1, Z2, Z3, Z4, and Z5 respectively. It can be seen that since the distribution of the sub-regions corresponding to the Z1 height is discrete and discontinuous, but in the component image corresponding to the Z1 height captured, these discontinuous image regions are all clear image regions, and only the coordinates of this part of the image region need to be recorded, that is, the coordinates of the distribution of the sub-regions at the Z1 height.

[0062] Furthermore, moving the image detection sensor along the optical axis direction of the image detection sensor and / or adjusting the shooting focal length of the image detection sensor according to the respective second height information of the sub-regions includes: excluding the sub-regions whose second height information exceeds the first threshold range.

[0063] Reference Figure 8 shown in, the second height information corresponding to the sub-regions where the sampling points 15, 16, and 22 are located is D2, and D2 is too high and exceeds the first threshold range. Then this sub-region can be excluded and the component image of this sub-region is not captured. This is because this sub-region may be too deeply sunken or too protruding and does not belong to the part of the panel surface that needs to be detected (it may be the position of the frame or edge perforation, etc.), so there is no need to capture a clear image for detection. Excluding such sub-regions can reduce the effective number of shootings and improve the efficiency.

[0064] Furthermore, excluding the sub-regions whose second height information exceeds the first threshold range includes: if the size of the sub-region whose second height information exceeds the first threshold range is smaller than the second threshold, it is determined that this sub-region is a defective region.

[0065] In the above example, reference Figure 9As shown, if the area formed by sampling points 15, 16, and 22 in the sub-region is too small and less than the second threshold, since there will be no protrusions or depressions in such a small range on the panel specimen, it can be determined that this sub-region is a defective area. For such defective areas with protrusions or depressions, there is no need to take pictures with the detection probe to confirm, which can improve the detection efficiency.

[0066] If the size of the sub-region where the second height information exceeds the first threshold range is greater than the third threshold, then it is determined that this sub-region is a characteristic structure area.

[0067] The characteristic structure area refers to some special characteristic structures existing on the panel surface, such as the sunken camera perforation area and the protruding installation and fixing area on the panel. For this type of characteristic structure area, there is no need to photograph and detect its defects.

[0068] Step S108: In the component images, take the area images of the corresponding sub-regions respectively and merge them into a detection area image for detection.

[0069] As in the above example, if the first component image of area 12347 and the second component image of area 5689 are taken, then the image contents of areas 1, 2, 3, 4, and 7 in the first component image and the image contents of areas 5, 6, 8, and 9 in the second component image can be taken for merging. Since these two parts of image contents are clear in their respective component images, the overall merged image is a clear image. Detecting on this merged clear image can improve the accuracy.

[0070] Furthermore, in this embodiment, the height measurement probe is preferably a linear height measurement probe. Refer to Figure 10 As shown, the height measurement probe, that is, the height detection probe 30, is replaced by a linear height measurement probe. The linear height measurement probe can perform height measurements on multiple sampling points on a line at one time. When the panel specimen slides under the linear height measurement probe, there is no need to stop, or it only needs to stop for a short time to complete the height measurement of all sampling points. Compared with the scheme of still needing to stagnate the panel specimen and using a single point-type height measurement probe to horizontally move and measure the height of sampling points, the efficiency is higher.

[0071] Furthermore, in one embodiment, as Figure 1 、 Figure 2 and Figure 10 shown, the position information of the sampling points on the panel surface plane can also be calibrated in advance by an optical alignment probe.

[0072] When the panel specimen is in an offset state in the horizontal XY plane, the corresponding offset of the position of the Mark on the panel specimen can be captured by two optical alignment probes 20, and compared with its position in the standard state. Then, the offset amount of the panel specimen (including angle deflection and position offset) can be obtained according to the coordinate system conversion relationship. Applying this offset amount to the positions of the sampling points can correct the positioning parameters of the sampling points.

[0073] In this embodiment, clustering the sampling points according to the corresponding first height information to obtain sub-regions related to height distribution within the detection field of view includes: setting the height interval for clustering according to the imaging range of the image detection sensor.

[0074] Reference Figure 11 shown Figure 11 As shown, four typical height difference patterns on the panel surface are presented to illustrate image fusion methods in various situations, but are not limited to the above four patterns. From left to right, they are the step difference pattern, the linear tilt pattern, the arc pattern, and the irregular step difference pattern. Among them, the step difference pattern represents the height difference pattern caused by product processes; the linear tilt pattern represents the angular monotonic height difference pattern caused by product warping or processes; the arc pattern represents the non-monotonic height difference pattern caused by abnormal protrusions / depressions or processes; the irregular step difference pattern represents the irregular step difference pattern caused by processes. Each height Z in the figure represents the second height information of the clustered sub-regions, and the corresponding thickness represents the imaging range of the image detection sensor, that is, the image detection sensor can clearly image within this thickness range.

[0075] For another example, if the images captured by the detection probe within the range of 10 microns above and below a specific height are images with acceptable clarity, clustering can be performed with 10 microns as an interval. If the first height information of 9 sampling points is as shown in Table 1 below:

[0076] Table 1

[0077] Sampling point number Detected first height information Clustering height (second height) 1 101 microns 100 microns 2 100 microns 100 microns 3 99 microns 100 microns 4 99 microns 100 microns 5 62 microns 60 microns 6 59 microns 60 microns 7 98 microns 100 microns 8 57 microns 60 microns 9 61 microns 60 microns

[0078] The first height information can be clustered into one type at sampling points between 95 and 105 microns, and the height information of the clustered sampling points can be normalized to the value of 100 microns. The first height information can be clustered into another type at sampling points between 55 and 65 microns, and the height information of the clustered sampling points can be normalized to the value of 60 microns. This enables, for the sub-regions where multiple sampling points of the same cluster are located, when the image detection sensor takes pictures, for the sampling points of the same cluster, only the sub-regions need to be adjusted, and the clarity of the captured images is acceptable. After clustering, the heights of multiple sampling points are normalized to the same height, that is, the second height information of the sub-region, and the image detection sensor only needs to adjust fewer times in the vertical direction to complete the shooting, thereby improving the efficiency.

[0079] To solve the above problems, the present invention also provides a panel detection device based on multi-depth height detection, as Figure 12 shown, including:

[0080] A height detection module 102, configured to obtain the first height information of sampling points within a detection field of view on the surface of the panel through a height measurement probe, and the number of sampling points included within the detection field of view is multiple.

[0081] A sampling point clustering module 104, configured to cluster the sampling points according to the corresponding first height information to obtain sub-regions related to height distribution within the detection field of view, and one sub-region information has second height information.

[0082] A component image acquisition module 106, configured to sequentially move the image detection sensor and / or adjust the shooting focal length of the image detection sensor along the optical axis direction of the image detection sensor according to the second height information corresponding to each sub-region, take pictures of the detection region, and obtain one or more component images corresponding to the sub-regions.

[0083] An image merging and detection module 108, configured to take the region images of the corresponding sub-regions in the component images and merge them into a detection region image for detection.

[0084] In one embodiment, the height measurement probe is a linear height measurement probe.

[0085] In one embodiment, the component image acquisition module 106 is further configured to exclude the sub-regions whose second height information exceeds the first threshold range.

[0086] In one embodiment, the component image acquisition module 106 is further configured to determine that the sub-region is a defective region when the size of the sub-region whose second height information exceeds the first threshold range is less than the second threshold.

[0087] In one embodiment, the sampling point clustering module 104 is configured to set a clustering height interval according to the imaging range of the image detection sensor.

[0088] Compared with the prior art, the panel detection method and device based on multi-depth altitude shooting of the present application have the following beneficial effects.

[0089] The panel detection method based on multi-depth altitude shooting of the present invention performs height sampling of multiple sampling points within the detection field of view of an image detection sensor (i.e., a detection probe), and clusters them into sub-regions of multiple heights within the detection field of view. When the image detection sensor can only move or focus in the direction perpendicular to the panel surface, it captures component images of multiple sub-regions, and then combines the component images into an image of the entire detection field of view. Since each component image is captured within the imaging range of the image detection sensor, the combined image of the entire detection field of view is clear in each region, thus ensuring the clarity of the image captured by the image detection sensor and improving the detection accuracy.

[0090] In this specification, the various embodiments are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same or similar parts among the various embodiments, reference can be made to each other.

[0091] The above are only the embodiments of the present invention, and do not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied to other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A panel detection method based on multi-depth high-altitude shooting, characterized in that Including: Obtaining first height information of sampling points within a detection field of view on the surface of the panel meter through a height measurement probe, where the number of sampling points included within the detection field of view is multiple; Clustering the sampling points according to the corresponding first height information to obtain sub-regions related to height distribution within the detection field of view, and one sub-region information has second height information; Sequentially moving the image detection sensor and / or adjusting the shooting focal length of the image detection sensor along the optical axis direction of the image detection sensor according to the second height information corresponding to each sub-region, taking pictures of the detection region, and obtaining one or more component images corresponding to the sub-regions; Taking the region images of the corresponding sub-regions in the component images respectively and merging them into a detection region image for detection.

2. The panel detection method based on multi-depth high-altitude shooting according to claim 1, wherein The height measurement probe is a linear height measurement probe.

3. The panel detection method based on multi-depth altitude shooting according to claim 1, characterized in that The sequentially moving the image detection sensor and / or adjusting the shooting focal length of the image detection sensor along the optical axis direction of the image detection sensor according to the second height information corresponding to each sub-region includes: Excluding the sub-regions where the second height information exceeds the first threshold range.

4. The panel detection method based on multi-depth altitude shooting according to claim 3, characterized in that, The excluding the sub-regions where the second height information exceeds the first threshold range includes: If the size of the sub-region where the second height information exceeds the first threshold range is less than the second threshold, determining that this sub-region is a defective region.

5. The panel detection method based on multi-depth high-altitude shooting according to claim 3, characterized in that, The excluding the sub-regions where the second height information exceeds the first threshold range includes: If the size of the sub-region where the second height information exceeds the first threshold range is greater than the third threshold, determining that this sub-region is a feature structure region.

6. The panel detection method based on multi-depth height detection shooting according to claim 1, characterized in that The clustering the sampling points according to the corresponding first height information to obtain sub-regions related to height distribution within the detection field of view includes: Setting a height interval for clustering according to the imaging range of the image detection sensor.

7. A panel detection device based on multi-depth high-altitude shooting, characterized in that, Including: A height detection module, configured to obtain first height information of sampling points within a detection field of view on the surface of the panel meter through a height measurement probe, where the number of sampling points included within the detection field of view is multiple; A sampling point clustering module, configured to cluster the sampling points according to the corresponding first height information to obtain sub-regions related to height distribution within the detection field of view, and one sub-region information has second height information; A component image acquisition module, configured to sequentially move the image detection sensor and / or adjust the shooting focal length of the image detection sensor along the optical axis direction of the image detection sensor according to the second height information corresponding to each sub-region, take pictures of the detection region, and obtain one or more component images corresponding to the sub-regions; An image merging and detection module, configured to take the region images of the corresponding sub-regions in the component images respectively and merge them into a detection region image for detection.

8. The panel detection device based on multi-depth high-altitude shooting according to claim 7, characterized in that, The height measurement probe is a linear height measurement probe.

9. The panel detection device based on multi-depth height detection shooting according to claim 7, characterized in that, The component image acquisition module is further configured to exclude the sub-regions where the second height information exceeds the first threshold range.

10. The panel detection device based on multi-depth height detection shooting according to claim 9, characterized in that, The component image acquisition module is further configured to determine that the sub-region is a defective region when the size of the sub-region where the second height information exceeds the first threshold range is less than the second threshold.

11. The panel detection device based on multi-depth height detection shooting according to claim 9, characterized in that, The component image acquisition module is further configured to determine that the sub-region is a feature structure region when the size of the sub-region where the second height information exceeds the first threshold range is greater than the third threshold.

12. The panel detection device based on multi-depth height detection shooting according to claim 7, characterized in that, The sampling point clustering module is configured to set a clustering height interval according to the imaging range of the image detection sensor.

Images