Defect detection method, product, device, system, and storage medium

Abstract

The application provides a defect detection method, a computer program product, a defect detection device, a detection system and a computer readable storage medium. The method comprises: acquiring first position information of a region of interest in a target object image; acquiring third position information of a suspected defect pixel in the target object image; determining a target region of interest according to the first position information and the third position information, wherein the target region of interest is a region of interest to which the suspected defect pixel belongs; and determining a defect detection parameter level corresponding to each target region of interest according to the number of suspected defect pixels in each target region of interest and the number of features contained in each target region of interest. In this way, the sensitivity of the defect detection parameter can be adaptively set based on the number of defects in the region of interest, thereby improving the detection efficiency while ensuring the detection accuracy.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of semiconductor detection, and in particular to a defect detection method, a computer program product, a defect detection device, a detection system and a computer readable storage medium. BACKGROUND

[0002] Before a workpiece is detected by a defect detection machine, detection parameters need to be set in advance, such as the relative movement speed of the workpiece and the detection device, and the exposure frequency of the detection device to obtain complete detection images. In addition, when the machine processes the detection images, the defect detection parameters need to be set in advance to obtain defects. How to set the defect detection parameters has become a problem to be solved. SUMMARY

[0003] To solve the above problems, the present application provides a defect detection method, a computer program product, a defect detection device, a detection system and a computer readable storage medium, which can adaptively set the sensitivity of the defect detection parameters based on the number of defects in the region of interest, and improve the detection efficiency under the premise of ensuring the detection accuracy.

[0004] To achieve the above purpose, the technical scheme of the embodiments of the present application is as follows:

[0005] In a first aspect, the embodiments of the present application provide a defect detection method, comprising:

[0006] obtaining first position information of a region of interest in a target object image;

[0007] obtaining third position information of suspected defect pixels in the target object image, determining a target region of interest according to the first position information and the third position information, the target region of interest being a region of interest to which the suspected defect pixels belong;

[0008] determining the level of defect detection parameters corresponding to each target region of interest according to the number of suspected defect pixels in each target region of interest and the number of features contained in each target region of interest.

[0009] In a second aspect, the embodiments of the present application provide a computer program product, and the instructions in the computer program product are executed by a processor of a defect detection device to make the defect detection device execute the defect detection method according to any of the embodiments of the present application.

[0010] In a third aspect, the embodiments of the present application provide a defect detection device, comprising a processor and a memory, the memory storing a computer program executable by the processor, and the computer program is executed by the processor to implement the defect detection method according to any of the embodiments of the present application.

[0011] In a fourth aspect, an embodiment of the present application provides a detection system, characterized in that the detection system comprises an image acquisition device and the defect detection device as described in any of the embodiments of the present application,

[0012] The image acquisition device is configured to rotate around the edge of the target object to sequentially acquire images of the edge part of the target object as the target object images and send the target object images to the defect detection device.

[0013] In a fifth aspect, an embodiment of the present application provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is executed by a controller to implement the defect detection method as described in any of the embodiments of the present application.

[0014] The defect detection method, the computer program product, the defect detection device, the detection system and the computer readable storage medium provided by the above embodiments of the present application can acquire the first position information of the region of interest in the target object image, and acquire the third position information of the suspected defect pixel in the target object image. According to the first position information and the third position information, the target region of interest to which the suspected defect pixel belongs can be accurately determined. According to the number of suspected defect pixels in each target region of interest and the number of features contained in each target region of interest, the defect detection parameter level corresponding to each target region of interest is determined. In this way, by determining the target region of interest to which the suspected defect pixel in each target sub-image belongs, the defect detection parameter level corresponding to each target region of interest is determined, so that the defect detection parameter level of each target region of interest can be adaptively set according to the characteristics of each target region of interest by simplifying the calculation, the detection precision is improved, the detection time is shortened, and the detection efficiency is improved. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 FIG. 1 shows an architecture schematic diagram of a detection system provided by an embodiment of the present application;

[0016] Figure 2 FIG. 2 shows a flow schematic diagram of a defect detection method provided by an embodiment of the present application;

[0017] Figure 3 FIG. 3 shows a schematic diagram of a target object image provided by an embodiment of the present application;

[0018] Figure 4 FIG. 4 shows a flow schematic diagram of another defect detection method provided by an embodiment of the present application;

[0019] Figure 5 FIG. 5 shows a schematic diagram of another target object image provided by an embodiment of the present application;

[0020] Figure 6A flowchart of another defect detection method provided by an embodiment of the present application is shown

[0021] Figure 7 A structure diagram of a defect detection device provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0022] The technical solutions of the present application will be further described in detail below in combination with the accompanying drawings and specific embodiments.

[0023] In order to make the purpose, technical solutions and advantages of the present application more clear, the present application will be further described in detail below in combination with the accompanying drawings, and the described embodiments should not be regarded as limiting the present application, all other embodiments obtained by those skilled in the art without making creative efforts fall within the scope of protection of the present application.

[0024] In the following description, the expression "some embodiments" describes a subset of all possible embodiments, and it should be noted that "some embodiments" can be the same subset or different subsets, and can be combined with each other without conflict.

[0025] In the following description, the terms "first, second, third" are only to distinguish similar objects, and do not represent a specific order of the objects. It can be understood that "first, second, third" can be interchanged in a specific order or sequence as allowed, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0026] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application.

[0027] The ROI (region of interest) can be a region that needs to be processed in a processed image, which is outlined by a regular pattern or an irregular image. One processed image can have multiple ROIs.

[0028] The die can be a chip on a processor, which is a core component of the processor. The chip manufacturing process is to divide a wafer into multiple units by a process flow, and to manufacture each unit into a die. However, various defects are inevitably generated in the chip manufacturing process, which are distributed on the wafer or the die. If a die has a defect, the die is a defective product and cannot be used to manufacture a processor. Therefore, the surface of the die or wafer can be detected for defects by optical detection and image processing technology.

[0029] However, in defect detection, fixed defect detection parameters are typically used to detect defects in the die image under test. But the feature complexity varies across different regions of the die image, resulting in some regions having flat patterns while others have complex patterns. Therefore, using the same defect detection parameter to process suspected defective areas in different regions may lead to false detections. For example, using a defect detection parameter that is sensitive to defects might detect correct defect pixels in flat areas, but in complex areas, it might detect normal pixels as defective pixels.

[0030] To address the aforementioned issues, the inventors proposed that, during defect detection, in order to confirm whether a defective pixel is a genuine defective pixel, corresponding defect detection parameters can be set based on the region of interest, thereby improving the accuracy of defective pixel detection.

[0031] In view of this, the inventors proposed to acquire multiple regions of interest in the target object image, determine which region of interest a suspected defect pixel belongs to based on its location, and determine the corresponding defect detection parameter level based on the characteristics of the target region of interest. In this way, the defect detection parameter level can be adaptively set, which is beneficial for subsequent defect detection based on the features of the target region of interest and improves the detection accuracy.

[0032] One aspect of this application provides a detection system. Figure 1 This diagram illustrates the architecture of a detection system provided in an embodiment of this application. Figure 1 As shown, the detection system may include an image acquisition device 11 and a defect detection device 12. The image acquisition device 11 may include an optical imaging sensor, such as an industrial camera, which can be used to rotate and scan around the edge of a target object, sequentially acquiring images of the edge portions of the target object and sending them as target object images to the defect detection device 12. The defect detection device 12, by executing the defect detection method provided in any embodiment of this application, determines the defect detection parameter level corresponding to each of the target regions of interest. Based on the determined defect detection parameter levels, the detection parameters of the defect detection machine on the workpiece can be adjusted accordingly. For example, the defect detection machine needs to pre-set defect detection thresholds and defect attention levels, so that detection can be performed on different regions of the target object image based on matching detection parameter levels, improving detection efficiency while ensuring detection accuracy.

[0033] Here, the defect detection device 12 can be various intelligent devices with storage and computing capabilities, such as a computer device, etc. The image acquisition device 11 can be a device with optical imaging capability, such as an industrial camera, etc. In the embodiments of the present application, the defect detection device 12 can be integrated with the image acquisition device 11, or independently arranged. The defect detection device 12 and the image acquisition device 11 can communicate with each other, for example, the defect detection device 12 can issue image acquisition instructions to the image acquisition device 11, and the image acquisition device 11 acquires images and sends the acquired images to the defect detection device after receiving the image acquisition instructions.

[0034] In an aspect of the embodiments of the present application, a defect detection method is provided, which can be applied to, for example, Figure 1 a defect detection device. Figure 2 A flowchart of a defect detection method provided by the embodiments of the present application is shown in FIG. 2, which can include but is not limited to S21, S22 and S23, and the details are as follows: Figure 2

[0035] S21, obtaining first position information of a region of interest in a target object image.

[0036] S22, obtaining third position information of a suspected defect pixel in the target object image, and determining a target region of interest according to the first position information and the third position information.

[0037] Here, the target region of interest is a region of interest to which the suspected defect pixel belongs.

[0038] S23, determining a defect detection parameter level corresponding to each target region of interest according to the number of suspected defect pixels in each target region of interest and the number of features contained in each target region of interest.

[0039] In the above embodiments, by determining the target region of interest to which the suspected defect pixel in the target image belongs, and determining the defect detection parameter level corresponding to each target region of interest, the defect detection parameter level of each target region of interest can be adaptively set according to the characteristics of each target region of interest by simplifying the calculation, so as to improve the detection accuracy, shorten the detection time and improve the detection efficiency.

[0040] The implementation of each step in the embodiments will be further described below. Figure 2

[0041] ​​In step S21, the target object image can be an image containing a target object. In the embodiments of the present application, the target object image can be an image corresponding to the local or global imaging of a chip. The region of interest can be an image region in the target object image, and can contain a region that needs to be processed. The region of interest can be regular or irregular in shape. The target object image can contain a plurality of regions of interest, and each region of interest can not overlap.

[0042] Figure 3 A schematic diagram of a target object image provided by an embodiment of the present application is shown. As shown in the figure, Figure 3 There are three regions of interest, ROI1, ROI2 and ROI3, on the target object image.

[0043] The first position information of the region of interest can be the position information of the region of interest in the target object image. The first position information of the region of interest can include coordinate information of the region of interest in the target object image. Optionally, the first position information of the region of interest can also include side length information of the region of interest.

[0044] It should be noted that the "coordinate information" involved in the embodiments of the present application can be coordinate information of a rectangular coordinate system. The X-axis is one coordinate axis of the rectangular coordinate system, and the Y-axis is another coordinate axis of the rectangular coordinate system, and the X-axis and the Y-axis are perpendicular to each other. For example, the rectangular coordinate can be a coordinate system constructed with the horizontal right as the X-axis and the vertical upward as the Y-axis. The coordinate information can include the coordinate along the X-axis and the coordinate along the Y-axis.

[0045] In the embodiments of the present application, the defect detection device can extract the regions of interest in the target object image by the image segmentation method, thereby obtaining a plurality of regions of interest. Different regions of interest contain different image features. The region of interest can be regular or irregular in shape.

[0046] In step S22, the suspected defect pixel can be a pixel corresponding to a suspected defect part in the target object image. The suspected defect pixel can be a true defect pixel or a false defect pixel. The target object image can include one or more suspected defect pixels. As shown in the figure, Figure 3 The suspected defect pixel is a black pixel point on the target object image.

[0047] The third position information of the suspected defect pixel can be the relative position information of the suspected defect pixel in the target object image. The third position information can include coordinate information of the suspected defect pixel.

[0048] In the embodiment of the present application, the third position information of each suspected defect pixel in the suspected defect region can be obtained by performing image recognition on the target object image to determine the suspected defect region in the target object image. According to the third position information of each suspected defect pixel and the first position information of the region of interest, the relationship between each suspected defect pixel and the region of interest is determined, and thus the target region of interest to which each suspected defect pixel belongs is determined. In this way, the region of interest containing the suspected defect pixel is determined by the position information, the range of the suspected defect pixel can be accurately determined, and the part not containing the suspected defect is filtered out, so as to reduce the subsequent processing process of the invalid image region and reduce the calculation amount.

[0049] It should be noted that in step S22, the target region of interest determination method is similar to the target region determination method in step S43 described below, and thus will not be described here.

[0050] In step S23, the number of suspected defect pixels in each target region of interest can be the number of suspected defect pixels contained in each target region of interest. The number of features contained in each target region of interest can be the number of pattern features contained in each target region of interest. The number of features can be used to represent the complexity of the image. When the number of features is larger, the image is more complex and the complexity is larger. When the number of features is smaller, the image is flatter and the complexity is smaller.

[0051] The defect detection parameter level can be the setting level of the defect detection parameter, which can be used to represent the sensitivity of the defect detection. The defect detection parameter level can be used to detect the real defect in the target region of interest. The defect detection parameter can be the defect detection threshold, the defect attention degree and other related detection parameters in the defect detection algorithm. Alternatively, the defect detection parameter level can include but is not limited to the insensitive level, the sensitive level and the super sensitive level. The insensitive level can represent the level of insensitivity to defect detection, the sensitive level represents the level of sensitivity to defect detection, and the super sensitive level represents the level of super sensitivity to defect detection.

[0052] Here, the defect detection parameter level can be inversely proportional to the number of features, and the defect detection parameter level can be proportional to the number of suspected defect pixels.

[0053] In this embodiment, an image feature extraction method can be used to determine the number of features contained in each target region of interest (ROI), and the number of suspected defect pixels within each ROI can be statistically analyzed to determine the total number of suspected defect pixels within each ROI. For each ROI, based on the number of features contained within the ROI and the number of suspected defect pixels, the corresponding defect detection parameter level is determined. In this way, the defect detection parameter level corresponding to each ROI can be adaptively adjusted according to its specific characteristics, so as to accurately detect the actual defects within each ROI and improve the defect detection accuracy.

[0054] Because the images of the dies to be tested have high magnification and large size, typically with a pixel size of 100,000 x 100,000 pixels or more, and the number of regions of interest (ROIs) in the images is large, reaching tens of thousands or more, the number of defect pixels to be tested is also relatively large, reaching tens of thousands or even millions. If each defect pixel is traversed through its ROI, the computational load is enormous, leading to prolonged defect detection time and reduced defect detection efficiency.

[0055] To address the aforementioned issues, the inventors proposed dividing the target image into multiple small sub-images. Based on the location of suspected defective pixels, the sub-image to which a pixel belongs is determined. Furthermore, based on the associated regions of interest (ROIs) within each sub-image, the target ROI to which the suspected defective pixel belongs is identified. Finally, based on the characteristics of the ROI, the corresponding defect detection parameter level is determined. This adaptive setting of the defect detection parameter level facilitates subsequent defect detection targeting the features of the ROI, improving detection accuracy. Moreover, by selecting the ROI from the target sub-images, the computational load during parameter setting is reduced, shortening defect detection time and improving defect detection efficiency.

[0056] In some embodiments, Figure 4 This document illustrates a flowchart of yet another defect detection method provided in an embodiment of this application, as shown below. Figure 4 As shown, the defect detection method is described in detail below:

[0057] In S22, before obtaining the third location information of suspected defect pixels in the target object image, and determining the target region of interest based on the first and third location information, the defect detection method further includes:

[0058] S41, divide the target object image into multiple sub-images, and obtain the second position information of the multiple sub-images;

[0059] S42, associate each region of interest with the plurality of sub-images based on the first position information of the region of interest and the second position information of the plurality of sub-images, to obtain an association relationship between the region of interest and the sub-image.

[0060] And S22 can be specifically implemented as the following steps:

[0061] S43, obtaining third position information of the suspected defect pixel in the target object image, and determining a target sub-image to which the suspected defect pixel belongs according to the third position information and the second position information respectively.

[0062] S44, determining a target region of interest in each target sub-image based on the association relationship.

[0063] In the above embodiment, the association relationship between the plurality of sub-images divided based on the target object image and the determined region of interest is determined, so as to determine the target sub-image to which the suspected defect pixel belongs, so as to filter out the part without suspected defect pixel from the target object image, reduce the calculation of the part without suspected defect pixel, and improve the detection efficiency. In addition, the target region of interest to which the suspected defect pixel belongs in each target sub-image is determined, and the corresponding defect detection parameter level of each target region of interest is determined, so that the calculation can be simplified, the characteristics of each target region of interest are realized, the defect detection parameter level of each target region of interest is adaptively set, the detection precision is improved, the detection time is shortened, and the detection efficiency is improved.

[0064] The implementation of each step in the embodiment will be further described below. Figure 4 The implementation of each step in the embodiment will be further described below.

[0065] Figure 5 Another schematic diagram of a target object image provided by the embodiment of the present application is shown. As shown in the figure, Figure 5 There are three regions of interest, ROI1, ROI2 and ROI3, on the target object image.

[0066] In step S41, the sub-image can be an image of a local region of the target object image. The size thereof can be 1 / n of the size of the target object image, and n can be a positive number. The plurality of sub-images can constitute a complete target object image, and each sub-image does not contain the same image region. Figure 5 As shown in the figure, there are four sub-images, block1, block2, block3 and block4, on the target object image.

[0067] The second position information of the sub-image can be the position information of the sub-image in the target object image. The second position information of the sub-image can include coordinate information of the sub-image.

[0068] In the embodiments of the present application, the defect detection device can divide the target object image into a plurality of image regions from left to right and from top to bottom according to the preset image scale or the preset number of sub-images, to obtain a plurality of sub-images.

[0069] It should be noted that some sub-images can include one or more regions of interest, and some sub-images can have no region of interest. In step S42, the association relationship can be an inclusion relationship between the sub-image and the region of interest, that is, the association relationship can indicate which regions of interest are included in a sub-image. It should be noted that the region of interest included in the sub-image can include a partial part of a region of interest, that is, a region of interest can span two or even more sub-images. As shown in FIG. 4, the association relationship between the region of interest and the sub-image is that block1 is associated with ROI3, block3 is associated with ROI3 and ROI1, and block3 is associated with ROI3 and ROI2. Figure 5

[0070] In the embodiments of the present application, the position relationship between each sub-image and each region of interest can be determined according to the second position information of the plurality of sub-images and the first position information of each region of interest, so as to associate each region of interest with each sub-image, to obtain the association relationship between the region of interest and the sub-image.

[0071] In step S43, the suspected defect pixel can be a pixel corresponding to a suspected defect part in the target object image. The suspected defect pixel can be a pixel of a true defect, or can be a pixel of a false defect. The target object image can include one or more suspected defect pixels. As shown in FIG. 5, the suspected defect pixel is a black pixel point on the target object image. Figure 5

[0072] The third position information of the suspected defect pixel can be information about the relative position of the suspected defect pixel in the target object image. The third position information can include coordinate information of the suspected defect pixel.

[0073] The target sub-image can refer to a sub-image involved by the suspected defect pixel in the target object image, that is, a sub-image containing the suspected defect pixel. One target sub-image can include one or more suspected defect pixels. As shown in FIG. 6, the target sub-image is block3. Figure 5

[0074] ​​​In the embodiment of the present application, the third position information of each suspected defect pixel in the suspected defect region can be obtained by performing image recognition on the target object image to determine the suspected defect region in the target object image. According to the third position information of the suspected defect pixel and the second position information of the plurality of sub-images, the position relationship between each suspected defect pixel and the plurality of sub-images is determined, so that the target sub-image to which each suspected defect pixel belongs is determined. In this way, the sub-image containing the suspected defect pixel is determined through the position information, which can narrow the range of the suspected defect pixel, facilitate subsequent searching for the corresponding region of interest, and filter out the part not containing the suspected defect, so as to reduce the subsequent processing process of the invalid image region and reduce the calculation amount.

[0075] In step S44, the target region of interest can be understood as the region of interest involved by the suspected defect pixel in the target object image, that is, the region of interest containing the suspected defect pixel. As shown in FIG. 4, the target sub-image is ROI2. Figure 5

[0076] In the embodiment of the present application, the target region of interest containing the suspected defect pixel in each target sub-image can be determined according to the association relationship between the region of interest and the sub-image.

[0077] In some embodiments, S44, determining the target region of interest in each target sub-image based on the association relationship can include:

[0078] For each target sub-image, determining the region of interest associated with the target sub-image according to the association relationship;

[0079] According to the first position information of the region of interest associated with the target sub-image and the third position information of the suspected defect pixel contained in the target sub-image, the target region of interest in the target sub-image is determined.

[0080] Here, the target region of interest can be the region of interest to which the suspected defect pixel belongs.

[0081] In the embodiment of the present application, the following steps can be performed for each target sub-image: determining the region of interest associated with the target sub-image according to the association relationship, and determining the position relationship between each region of interest and the suspected defect pixel in the target sub-image according to the first position information of the region of interest associated with the target sub-image and the third position information of the suspected defect pixel contained in the target sub-image, and finally determining the target region of interest containing the suspected defect pixel in the target sub-image. In this way, the target region of interest in each target sub-image can be determined.

[0082] ​In the above embodiment, the target sub-image is associated with a region of interest, and the target region of interest is determined according to the position information of the associated region of interest and the suspected pixel, so that the region of interest containing the suspected defect pixel can be found through a small amount of calculation, the searching efficiency is improved, and the calculation amount is reduced.

[0083] In some embodiments, the target region of interest in the target sub-image is determined according to the first position information of the region of interest associated with the target sub-image and the third position information of the suspected defect pixel contained in the target sub-image, and includes:

[0084] According to the first position information of the region of interest associated with the target sub-image, the X-axis coordinate of the target reference point in the region of interest is sorted to form a first array, and the Y-axis coordinate of the target reference point is sorted to form a second array.

[0085] According to the third position information of the suspected defect pixel contained in the target sub-image, a first target region of interest sequence adjacent to the suspected defect pixel in the first array and a second target region of interest sequence adjacent to the suspected defect pixel in the second array are determined.

[0086] The region of interest that overlaps in the first target region of interest sequence and the second target region of interest sequence is determined as the target region of interest in the target sub-image.

[0087] Here, the region of interest associated with the target sub-image can be multiple. The first position information can include coordinate information. The target reference point can be a point representing the range of the region of interest. Alternatively, the target reference point can be the point in the region of interest that is the largest or smallest along the coordinate axis in the rectangular coordinate system. For example, the target reference point can be any one corner point of the region of interest. In the case of a rectangular region of interest, the target reference point can include at least one of the upper left corner point, the lower left corner point, the lower right corner point, and the upper right corner point of the region of interest.

[0088] The first array can be a permutation array of all regions of interest in a target sub-image, and the first array is sorted according to the size of the X-axis coordinate of the target reference point of each region of interest. The second array can be a permutation array of all regions of interest in a target sub-image, and the second array is sorted according to the size of the Y-axis coordinate of the target reference point of each region of interest.

[0089] In the embodiments of the present application, the first position information of all the regions of interest associated with the target sub-image can be used to sort all the regions of interest in the target sub-image according to the X-axis coordinate of the respective target reference point to form a first array, and according to the Y-axis coordinate of the respective target reference point to form a second array. Here, the sorting method can be from small to large or from large to small.

[0090] The first target region of interest sequence can be a sequence formed by part of the elements in the first array. In the embodiments of the present application, the X-axis coordinate of the target reference point of the region of interest in the first array can be compared with the X-axis coordinate of the suspected defective pixel to determine the region of interest adjacent to the suspected defective pixel from the first array, thereby forming the first target region of interest sequence. The region of interest adjacent to the suspected defective pixel can be a region of interest that is more likely to contain the suspected defective pixel.

[0091] The second target region of interest sequence can be a sequence formed by part of the elements in the second array. In the embodiments of the present application, the Y-axis coordinate of the target reference point of the region of interest in the second array can be compared with the Y-axis coordinate of the suspected defective pixel to determine the region of interest adjacent to the suspected defective pixel from the second array, thereby forming the second target region of interest sequence.

[0092] For example, in the case where the X-axis is positive in the horizontal right direction, the Y-axis is positive in the vertical upward direction, and the target reference point is the left upper corner point, the region of interest whose X-axis coordinate of the target reference point is smaller than the X-axis coordinate of the suspected defective pixel can be regarded as the region of interest adjacent to the suspected defective pixel in the first array, thereby forming the first target region of interest sequence. The region of interest whose Y-axis coordinate of the target reference point is greater than the Y-axis coordinate of the suspected defective pixel can be regarded as the region of interest adjacent to the suspected defective pixel in the second array, thereby forming the second target region of interest sequence.

[0093] For example, in the case where the X-axis is positive in the horizontal right direction, the Y-axis is positive in the vertical upward direction, and the target reference point is the left lower corner point, the region of interest whose X-axis coordinate of the target reference point is smaller than the X-axis coordinate of the suspected defective pixel can be regarded as the region of interest adjacent to the suspected defective pixel in the first array, thereby forming the first target region of interest sequence. The region of interest whose Y-axis coordinate of the target reference point is smaller than the Y-axis coordinate of the suspected defective pixel can be regarded as the region of interest adjacent to the suspected defective pixel in the second array, thereby forming the second target region of interest sequence.

[0094] For example, in the case that the X-axis is horizontally right as a positive direction, the Y-axis is vertically upward as a positive direction, and the target reference point is the right upper corner point, the interested region with the X-axis coordinate of the target reference point greater than the X-axis coordinate of the suspected defect pixel can be regarded as the interested region adjacent to the suspected defect pixel in the first array, to form the first target interested region sequence. The interested region with the Y-axis coordinate of the target reference point greater than the Y-axis coordinate of the suspected defect pixel can be regarded as the interested region adjacent to the suspected defect pixel in the second array, to form the second target interested region sequence.

[0095] For example, in the case that the X-axis is horizontally right as a positive direction, the Y-axis is vertically upward as a positive direction, and the target reference point is the right upper corner point, the interested region with the X-axis coordinate of the target reference point greater than the X-axis coordinate of the suspected defect pixel can be regarded as the interested region adjacent to the suspected defect pixel in the first array, to form the first target interested region sequence. The interested region with the Y-axis coordinate of the target reference point greater than the Y-axis coordinate of the suspected defect pixel can be regarded as the interested region adjacent to the suspected defect pixel in the second array, to form the second target interested region sequence.

[0096] It should be noted that the way of judging the interested region adjacent to the suspected defect pixel in the first array and the second array is adaptively modified according to the direction of the coordinate axis of the constructed rectangular coordinate and the selected target reference point.

[0097] In the embodiment of the present application, the first target interested region sequence composed of the interested regions with a higher possibility of containing the suspected defect pixel is judged from the X-axis coordinate, and the second target interested region sequence composed of the interested regions with a higher possibility of containing the suspected defect pixel is judged from the Y-axis coordinate. Therefore, the interested region overlapping in the first target interested region sequence and the second target interested region sequence can be found, and the overlapping interested region can be determined as the target interested region containing the suspected defect pixel in the target sub-image. In this way, the target interested region in the target sub-image is jointly judged from different dimensions.

[0098] In the above embodiment, two sequences of the interested regions containing the suspected defect pixel are determined by respectively comparing the different coordinate axis coordinate values of the target reference point of the interested region and the suspected defect pixel, and the overlapping interested region in the two sequences is determined as the target interested region. In this way, the target interested region in the target sub-image is jointly judged from different dimensions, so that the target interested region is found by traversal, the operation amount is reduced, and the efficiency of finding the target interested region is improved.

[0099] It should be noted that in the case where the number of regions of interest associated with the target sub-image is 1, the sorting step can be omitted, and it is directly determined whether the region of interest is adjacent to the suspected defect pixel.

[0100] In some embodiments, according to the third position information of the suspected defect pixel contained in the target sub-image, determining the first target region of interest sequence adjacent to the suspected defect pixel in the first array and the second target region of interest sequence adjacent to the suspected defect pixel in the second array can include:

[0101] According to the X-axis coordinate of the suspected defect pixel, searching for a first reference array element in the first array, and determining a part before or after the first reference array element in the first array as the first target region of interest sequence adjacent to the suspected defect pixel;

[0102] According to the Y-axis coordinate of the suspected defect pixel, searching for a second reference array element in the second array, and determining a part before or after the second reference array element in the second array as the second target region of interest sequence adjacent to the suspected defect pixel.

[0103] Here, the first reference array element can be any one of the two adjacent array elements in the first array. Among them, the X-axis coordinates of the target reference points corresponding to the two adjacent array elements in the first array are opposite in size to the X-axis coordinate of the suspected defect pixel. For example, the X-axis coordinate of the target reference point corresponding to the previous array element is smaller than the X-axis coordinate of the suspected defect pixel, and the X-axis coordinate of the target reference point corresponding to the next array element is greater than the X-axis coordinate of the suspected defect pixel.

[0104] The part before or after the first reference array element in the first array can refer to the part of the array before the first reference array element, including the first reference array element, or the part of the array after the first reference array element, including the first reference array element.

[0105] The second reference array element can be any one of the two adjacent array elements in the second array. Among them, the Y-axis coordinates of the target reference points corresponding to the two adjacent array elements in the second array are opposite in size to the Y-axis coordinate of the suspected defect pixel. For example, the Y-axis coordinate of the target reference point corresponding to the previous array element is smaller than the Y-axis coordinate of the suspected defect pixel, and the Y-axis coordinate of the target reference point corresponding to the next array element is greater than the Y-axis coordinate of the suspected defect pixel.

[0106] The part before or after the second reference array element in the second array can refer to the part of the array before the second reference array element, including the second reference array element, or the part of the array after the second reference array element, including the second reference array element.

[0107] It should be noted that which part of the array is selected as the target region of interest sequence needs to be determined according to the sorting manner of the elements in the array and the target reference point.

[0108] For example, in the case of the target reference point being the upper left corner point, the elements in the first array are arranged in ascending order, the upper left corner point X-axis coordinate in the Ath array element in the first array is not greater than the X-axis coordinate of the suspected defect pixel, and the upper left corner point X-axis coordinate in the A+1th array element is greater than the X-axis coordinate of the suspected defect pixel. It can be determined that the Ath array element is the first reference array element in the first array, and the first target region of interest sequence is formed by the first array element to the Ath array element in the first array.

[0109] The elements in the second array are arranged in ascending order, the upper left corner point Y-axis coordinate in the B-1th array element in the second array is not greater than the Y-axis coordinate of the suspected defect pixel, and the upper left corner point Y-axis coordinate in the Bth array element is greater than the Y-axis coordinate of the suspected defect pixel. It can be determined that the Bth array element is the second reference array element in the second array, and the second target region of interest sequence is formed by the Bth array element to the last array element in the second array.

[0110] In the case of the target reference point being the upper right corner point, the elements in the first array are arranged in ascending order, the upper right corner point X-axis coordinate in the C-1th array element in the first array is not greater than the X-axis coordinate of the suspected defect pixel, and the upper right corner point X-axis coordinate in the Cth array element is greater than the X-axis coordinate of the suspected defect pixel. It can be determined that the Cth array element is the first reference array element in the first array, and the first target region of interest sequence is formed by the Cth array element to the last array element in the first array.

[0111] The elements in the second array are arranged in ascending order, the upper right corner point Y-axis coordinate in the D-1th array element in the second array is not greater than the Y-axis coordinate of the suspected defect pixel, and the upper left corner point Y-axis coordinate in the Dth array element is greater than the Y-axis coordinate of the suspected defect pixel. It can be determined that the Dth array element is the second reference array element in the second array, and the second target region of interest sequence is formed by the Dth array element to the last array element in the second array.

[0112] In the embodiments of the present application, the coordinates of the suspected defective pixel and the target reference point are compared, and a first reference array element is selected from a first array, and a second reference array element is selected from a second array. The array elements before or after the first reference array element in the first array and the first reference array element form a first target region of interest sequence adjacent to the suspected defective pixel in the first array, and the array elements before or after the second reference array element in the second array and the second reference array element form a second target region of interest sequence adjacent to the suspected defective pixel in the second array.

[0113] In the above embodiments, the reference array elements in each array are found to determine each target region of interest sequence, which is beneficial to quickly determine each target region of interest sequence from each array.

[0114] In some embodiments, according to the first position information of the target sub-image associated region of interest, the X-axis coordinate size of the target reference point in the region of interest is sorted to form a first array, and the Y-axis coordinate size of the target reference point is sorted to form a second array, which can include:

[0115] The target reference point includes a first corner point of the region of interest. According to the first position information of the target sub-image associated region of interest, the X-axis coordinate size of the first corner point of the region of interest is sorted to form a first array corresponding to the first corner point, and the Y-axis coordinate size of the first corner point of the region of interest is sorted to form a second array corresponding to the first corner point.

[0116] Here, the first position information can include coordinate information. The region of interest can be a rectangle, and the target reference point can include a first corner point of the region of interest. The first corner point can be any corner point of the region of interest. Alternatively, the first corner point can be a top-left corner point, a bottom-left corner point, a top-right corner point, or a bottom-right corner point.

[0117] The first array can be an array formed by sorting each region of interest in a target sub-image according to the X-axis coordinate size of the respective first corner point. The second array can be an array formed by sorting each region of interest in a target sub-image according to the Y-axis coordinate size of the respective first corner point.

[0118] In the embodiments of the present application, according to the first position information of all target sub-image associated regions of interest, the X-axis coordinate size of the first corner point of all regions of interest in the target sub-image is sorted to form a first array, and the Y-axis coordinate size of the respective target reference point is sorted to form a second array.

[0119] In the embodiments of the present application, the first target region of interest sequence can be formed by comparing the X-axis coordinate of the first corner point of the region of interest in the first array with the X-axis coordinate of the suspected defect pixel, and determining the region of interest adjacent to the suspected defect pixel from the first array. The first target region of interest sequence can also be formed by comparing the Y-axis coordinate of the first corner point of the region of interest in the second array with the Y-axis coordinate of the suspected defect pixel, and determining the region of interest adjacent to the suspected defect pixel from the second array. The target region of interest containing the suspected defect pixel in the target sub-image can be determined by finding the overlapping region of interest in the first target region of interest sequence and the second target region of interest sequence.

[0120] In the above embodiments, one corner point of the region of interest is taken as the target reference point, and the two arrays are formed according to the coordinate size of the corner point, which is beneficial to the subsequent accurate determination of the region of interest where the suspected defect pixel is located.

[0121] In some embodiments, according to the first position information of the region of interest associated with the target sub-image, the first array is formed based on the X-axis coordinate size of the target reference point in the region of interest, and the second array is formed based on the Y-axis coordinate size of the target reference point, including:

[0122] The target reference point includes the first corner point and the second corner point of the region of interest, wherein the second corner point is the diagonal of the first corner point. According to the first position information of the region of interest associated with the target sub-image, the first array corresponding to the first corner point is formed based on the X-axis coordinate size of the first corner point of the region of interest, the first array corresponding to the second corner point is formed based on the X-axis coordinate size of the second corner point of the region of interest, the second array corresponding to the first corner point is formed based on the Y-axis coordinate size of the first corner point of the region of interest, and the second array corresponding to the second corner point is formed based on the Y-axis coordinate size of the second corner point of the region of interest.

[0123] Here, the first position information can include coordinate information. The region of interest can be a rectangle, and the target reference point can include the first corner point of the region of interest and the first corner point of the region of interest. The second corner point is the diagonal point of the first corner point, that is, the first corner point and the second corner point are a pair of diagonal points.

[0124] For example, in the case where the first corner point is the upper left corner, the second corner point is the lower right corner. In the case where the first corner point is the lower left corner, the second corner point is the upper right corner. In the case where the first corner point is the lower right corner, the second corner point is the upper left corner. In the case where the first corner point is the upper right corner, the second corner point is the lower left corner.

[0125] The first array can include a first array corresponding to each of the two opposite corners, that is, can include a first array corresponding to the first corner and a first array corresponding to the second corner. The first array corresponding to the first corner is an arrangement array formed by sorting each region of interest in the target sub-image according to the X-axis coordinate size of the first corner thereof. The first array corresponding to the second corner is an arrangement array formed by sorting each region of interest in the target sub-image according to the X-axis coordinate size of the second corner thereof.

[0126] The second array can include a second array corresponding to each of the two opposite corners, that is, can include a second array corresponding to the first corner and a second array corresponding to the second corner. The second array corresponding to the first corner is an arrangement array formed by sorting each region of interest in the target sub-image according to the Y-axis coordinate size of the first corner thereof. The second array corresponding to the second corner is an arrangement array formed by sorting each region of interest in the target sub-image according to the Y-axis coordinate size of the second corner thereof.

[0127] In the embodiment of the present application, by comparing the X-axis coordinate of the first corner of the region of interest in the first array corresponding to the first corner with the X-axis coordinate of the suspected defect pixel, the region of interest adjacent to the suspected defect pixel can be determined from the first array, and a first target region of interest sequence corresponding to the first corner is formed.

[0128] In addition, by comparing the X-axis coordinate of the second corner of the region of interest in the first array corresponding to the second corner with the X-axis coordinate of the suspected defect pixel, the region of interest adjacent to the suspected defect pixel can be determined from the first array, and a first target region of interest sequence corresponding to the second corner is formed.

[0129] In addition, by comparing the Y-axis coordinate of the first corner of the region of interest in the second array corresponding to the first corner with the Y-axis coordinate of the suspected defect pixel, the region of interest adjacent to the suspected defect pixel can be determined from the second array, and a second target region of interest sequence corresponding to the first corner is formed.

[0130] In addition, by comparing the Y-axis coordinate of the second corner of the region of interest in the second array corresponding to the second corner with the Y-axis coordinate of the suspected defect pixel, the region of interest adjacent to the suspected defect pixel can be determined from the second array, and a second target region of interest sequence corresponding to the second corner is formed.

[0131] In the embodiments of the present application, the coincident regions of interest can be found in the first target region of interest series corresponding to the first corner point, the first target region of interest series corresponding to the second corner point, the second target region of interest series corresponding to the first pair of corner points, and the second target region of interest series corresponding to the second pair of corner points, and the coincident regions of interest are determined as the target regions of interest containing the suspected defective pixels in the target sub-image, so that the target regions of interest are more accurately determined through multiple comparisons.

[0132] In the above embodiments, the corner points of the regions of interest opposite to each other are taken as the target reference points, and the corner points are sorted according to the coordinate size to form four groups of arrays, which facilitates subsequent more rapid and accurate determination of the target regions of interest.

[0133] In some embodiments, after S21, each region of interest is numbered. In this way, it is beneficial to quickly find each region of interest.

[0134] In some embodiments, according to the first position information of the regions of interest associated with the target sub-image, the first corner point of the region of interest is sorted according to the X-axis coordinate size to form a first array corresponding to the first corner point, and the first corner point of the region of interest is sorted according to the Y-axis coordinate size to form a second array corresponding to the first corner point, including:

[0135] According to the first position information of the regions of interest associated with the target sub-image, the first corner point of the region of interest is sorted according to the X-axis coordinate size in ascending order or descending order, and one or more of the number of each region of interest, the coordinate value of the first corner point, and the length and width of the region of interest are taken as the array elements corresponding thereto to form a first array corresponding to the first corner point;

[0136] The first corner point of the region of interest is sorted according to the Y-axis coordinate in ascending order or descending order, and one or more of the number of each region of interest, the coordinate value of the first corner point, and the length and width of the region of interest are taken as the array elements corresponding thereto to form a second array corresponding to the first corner point.

[0137] Here, the first corner point of the region of interest can be any corner point of the region of interest. The region of interest can be a rectangle. The first position information of the region of interest can include one or more of the number of the region of interest, the coordinate value of the corner point, and the length and width of the region of interest. Ascending order can refer to arranging in order from small to large, and descending order can refer to arranging in order from large to small.

[0138] The first array corresponding to the first corner point is formed according to the respective first corner point of each region of interest in the target sub-image in ascending order or descending order according to the X-axis coordinate size. Each array element in the first array corresponds to a region of interest in the target sub-image. One array element in the first array can include one or more of the number of the region of interest, the first corner point coordinate value of the region of interest, the length and width of the region of interest.

[0139] The second array corresponding to the first corner point is formed according to the respective first corner point of each region of interest in the target sub-image in ascending order or descending order according to the Y-axis coordinate size. Each array element in the second array corresponds to a region of interest in the target sub-image. One array element in the second array can include one or more of the number of the region of interest, the first corner point coordinate value of the region of interest, the length and width of the region of interest.

[0140] In the embodiments of the present application, the number of the region of interest can be used to identify the region of interest. The length and width of the region of interest can be used to determine the image region size of the region of interest. Based on the first corner point coordinate value of the region of interest and the length and width of the region of interest, the position and image region size of the region of interest in the target object image can be accurately determined.

[0141] In the above embodiments, one or more of the number, one corner point coordinate, and length and width are used as an array element. In this way, the corresponding region of interest of the array element can be quickly found by the array element serial number, and the range of the region of interest corresponding to one array element can be accurately determined, which is convenient for subsequent operation.

[0142] In some embodiments, the first array corresponding to the second corner point is formed based on the X-axis coordinate size of the second corner point of the region of interest, and the second array corresponding to the second corner point is formed based on the Y-axis coordinate size of the second corner point of the region of interest, including:

[0143] The first array corresponding to the second corner point is formed based on the X-axis coordinate size of the second corner point of the region of interest in ascending order or descending order according to the X-axis coordinate from small to large, and one or more of the number of each region of interest, the second corner point coordinate value, and the length and width of the region of interest is used as the array element corresponding thereto.

[0144] The second array corresponding to the second corner point is formed based on the Y-axis coordinate size of the second corner point of the region of interest in ascending order or descending order according to the Y-axis coordinate, and one or more of the number of each region of interest, the second corner point coordinate value, and the length and width of the region of interest is used as the array element corresponding thereto.

[0145] Here, the second corner point of the region of interest can be a diagonal corner point of the first corner point. The region of interest can be a rectangle. The first position information of the region of interest can include one or more of a number of the region of interest, coordinate values of the corner points, a length and a width of the region of interest. The ascending order arrangement can refer to arranging in order from small to large, and the descending order arrangement can refer to arranging in order from large to small.

[0146] The first array corresponding to the second corner point is an array formed by arranging the regions of interest in the target sub-image in ascending or descending order according to the X-axis coordinate values of the respective second corner points. Each array element in the first array corresponds to a region of interest in the target sub-image. One array element in the first array can include one or more of a number of the region of interest, coordinate values of the first corner point of the region of interest, a length and a width of the region of interest.

[0147] The second array corresponding to the second corner point is an array formed by arranging the regions of interest in the target sub-image in ascending or descending order according to the Y-axis coordinate values of the respective second corner points. Each array element in the second array corresponds to a region of interest in the target sub-image. One array element in the second array can include one or more of a number of the region of interest, coordinate values of the first corner point of the region of interest, a length and a width of the region of interest.

[0148] In the embodiments of the present application, the number of the region of interest can be used to identify the region of interest. The length and the width of the region of interest can be used to determine the image region size of the region of interest. Based on the second corner point coordinate values of the region of interest and the length and the width of the region of interest, the position and the image region size of the region of interest in the target object image can be determined.

[0149] In the above embodiments, one or more of the number, one corner point coordinate, and the length and the width are taken as an array element. In this way, the corresponding region of interest of the array element can be quickly found by the array element serial number, and the range of the region of interest corresponding to one array element can be accurately determined, which is convenient for subsequent operations.

[0150] In some embodiments, according to the third position information of the suspected defect pixel included in the target sub-image, a first target region of interest sequence adjacent to the suspected defect pixel in the first array and a second target region of interest sequence adjacent to the suspected defect pixel in the second array are determined, including:

[0151] According to the third position information of the suspected defect pixel included in the target sub-image;

[0152] determining the first target region of interest sequence adjacent to the suspected defective pixel in the first array based on the size of the X coordinate of the suspected defective pixel and the X coordinate of the middle array element of the first array, and taking the array element in front of or behind the first array as a comparison object;

[0153] determining the second target region of interest sequence adjacent to the suspected defective pixel in the second array based on the size of the Y coordinate of the suspected defective pixel and the X coordinate of the middle array element of the second array, and taking the array element in front of or behind the second array as a comparison object.

[0154] Here, the third position information can include coordinate information of the suspected defective pixel. The array element in front of or behind the first array can be an array element in front of the middle array element in the first array, or an array element behind the middle array element in the first array. The array element in front of or behind the second array can be an array element in front of the middle array element in the second array, or an array element behind the middle array element in the first array.

[0155] In the embodiment of the present application, the middle array element of the first array is selected, the size of the X coordinate of the target reference point in the middle array element and the X coordinate of the suspected defective pixel are compared, when the X coordinate of the target reference point in the middle array element is less than the X coordinate of the suspected defective pixel, the array element with the X coordinate of the target reference point greater than the X coordinate of the target reference point in the middle array element is selected as a comparison object, and the first target region of interest sequence adjacent to the suspected defective pixel is determined from the first array.

[0156] When the X coordinate of the target reference point in the middle array element is greater than the X coordinate of the suspected defective pixel, the array element with the X coordinate of the target reference point less than the X coordinate of the target reference point in the middle array element is selected as a comparison object, and the first target region of interest sequence adjacent to the suspected defective pixel is determined from the first array.

[0157] It should be noted that when the array elements in the first array are arranged in ascending order, the array element with the X coordinate of the target reference point greater than the X coordinate of the target reference point in the middle array element can be an array element behind the middle array element in the first array. The array element with the X coordinate of the target reference point less than the X coordinate of the target reference point in the middle array element can be an array element in front of the middle array element in the first array.

[0158] In the case that the array elements in the first array are arranged in descending order, the array element whose X-axis coordinate of the target reference point is greater than the X-axis coordinate of the target reference point of the middle array element can be the array element in front of the middle array element in the first array. The array element whose X-axis coordinate of the target reference point is less than the X-axis coordinate of the target reference point of the middle array element can be the array element behind the middle array element in the first array.

[0159] The middle array element of the second array is selected, and the Y-axis coordinate of the target reference point in the middle array element is compared with the axis coordinate of the suspected defective pixel. In the case that the Y-axis coordinate of the target reference point in the middle array element is less than the axis coordinate of the suspected defective pixel, the array element whose Y-axis coordinate of the target reference point is greater than the Y-axis coordinate of the target reference point of the middle array element is selected as the comparison object, and the second target region of interest sequence adjacent to the suspected defective pixel is determined from the second array.

[0160] In the case that the Y-axis coordinate of the target reference point in the middle array element is greater than the axis coordinate of the suspected defective pixel, the array element whose Y-axis coordinate of the target reference point is less than the Y-axis coordinate of the target reference point of the middle array element is selected as the comparison object, and the second target region of interest sequence adjacent to the suspected defective pixel is determined from the second array.

[0161] It should be noted that in the case that the array elements in the first array are arranged in ascending order, the array element whose Y-axis coordinate of the target reference point is greater than the Y-axis coordinate of the target reference point of the middle array element can be the array element behind the middle array element in the first array. The array element whose Y-axis coordinate of the target reference point is less than the Y-axis coordinate of the target reference point of the middle array element can be the array element in front of the middle array element in the first array.

[0162] In the case that the array elements in the first array are arranged in descending order, the array element whose Y-axis coordinate of the target reference point is greater than the Y-axis coordinate of the target reference point of the middle array element can be the array element in front of the middle array element in the first array. The array element whose Y-axis coordinate of the target reference point is less than the Y-axis coordinate of the target reference point of the middle array element can be the array element behind the middle array element in the first array.

[0163] In the above embodiment, by comparing the coordinate values of the middle array element and the suspected defective pixel, the array element in front of or behind the array is determined as the comparison object, and the target region of interest sequence adjacent to the suspected defective pixel is determined from the array. In this way, the target region of interest sequence is quickly determined by the bisection method, the calculation amount of the comparison operation is reduced, and the formation speed of the target region of interest sequence is improved.

[0164] In some embodiments, according to the third position information of the suspected defective pixel contained in the target sub-image, the first target region of interest number series adjacent to the suspected defective pixel in the first array and the second target region of interest number series adjacent to the suspected defective pixel in the second array are determined, including:

[0165] According to the third position information of the suspected defective pixel contained in the target sub-image;

[0166] Determining the middle array element of the first array;

[0167] Based on the size of the X coordinate of the suspected defective pixel and the X axis coordinate of the middle array element of the first array, taking the array element in front of or behind the first array as a comparison array, returning the middle array element of the determined comparison array, based on the size of the X coordinate of the suspected defective pixel and the X axis coordinate of the middle array element of the comparison array, taking the array element in front of or behind the comparison array again as another comparison array, until there are two array elements in the comparison array, taking the array elements in the comparison array as comparison objects to determine the first target region of interest number series adjacent to the suspected defective pixel in the first array;

[0168] Determining the middle array element of the second array;

[0169] Based on the size of the Y coordinate of the suspected defective pixel and the Y axis coordinate of the middle array element of the second array, taking the array element in front of or behind the second array as a comparison array, returning the middle array element of the determined comparison array, based on the size of the Y coordinate of the suspected defective pixel and the Y axis coordinate of the middle array element of the comparison array, taking the array element in front of or behind the comparison array again as another comparison array, until there are two array elements in the comparison array, taking the array elements in the comparison array as comparison objects to determine the second target region of interest number series adjacent to the suspected defective pixel in the second array.

[0170] Here, by using the bisection method multiple times to determine the comparison array with two array elements in the first array element and the second array element, and taking the array elements in the comparison array as comparison objects, the first target region of interest number series adjacent to the suspected defective pixel in the first array and the second target region of interest number series adjacent to the suspected defective pixel in the second array are determined respectively.

[0171] It should be noted that the steps of determining the first target region of interest number series and determining the second target region of interest number series are the same as in the foregoing embodiments, which will not be repeated here.

[0172] In the above embodiments, the target region of interest number series is determined by using the bisection method multiple times, which can further reduce the calculation amount of the comparison operation and further improve the efficiency of forming the target region of interest number series.

[0173] In some embodiments, the overlapped region of interest in the first target region of interest series and the second target region of interest series is determined as the target region of interest in the target sub-image, comprising:

[0174] Selecting the series containing the least number of regions of interest from the first target region of interest series and the second target region of interest series as a reference series;

[0175] In sequence, taking the region of interest contained in the reference series as the object, judging whether the other target region of interest series includes the overlapped region of interest, until the overlapped region of interest in the first target region of interest series and the second target region of interest series is found, and determining the overlapped region of interest as the target region of interest in the target sub-image.

[0176] Here, the reference series is the series containing the least number of regions of interest in the first target region of interest series and the second target region of interest series, that is, the reference series can be the first target region of interest series or the second target region of interest series. The other target region of interest series can be the series other than the reference series in the first target region of interest series and the second target region of interest series. The overlapped region of interest can be the same region of interest as the region of interest contained in the reference series. The overlapped region of interest in the first target region of interest series and the second target region of interest series.

[0177] In sequence, taking the region of interest contained in the reference series as the object, judging whether the other target region of interest series includes the overlapped region of interest, can be understood as taking the region of interest contained in the reference series as the object in sequence, judging whether the other target region of interest series includes the overlapped region of interest of the object. That is, traversing the other target region of interest series in sequence with each region of interest contained in the reference series to find whether the other target region of interest series includes the overlapped region of interest.

[0178] The target region of interest can be the region of interest to which the suspected defect pixel contained in the target sub-image belongs.

[0179] In the embodiments of the present application, the series containing the least number of regions of interest is selected from the first target region of interest series and the second target region of interest series as the reference series, each region of interest in the reference series is traversed in sequence to find the overlapped region of interest in the first target region of interest series and the second target region of interest series, and the overlapped region of interest is taken as the target region of interest in the target sub-image.

[0180] In the above embodiment, the number series with the least number of regions of interest is taken as the reference number series, and other number series are traversed to find the target region of interest. In this way, the target region of interest can be found with the least amount of calculation, the search time is shortened, the search efficiency is improved, and the defect detection time is shortened.

[0181] In some embodiments, in a case where the first target region of interest number series includes a first target region of interest number series corresponding to the first diagonal point and a first target region of interest number series corresponding to the second diagonal point, and the second target region of interest number series includes a second target region of interest number series corresponding to the first diagonal point and a second target region of interest number series corresponding to the second diagonal point, the regions of interest that overlap in the first target region of interest number series and the second target region of interest number series are determined as the target region of interest in the target sub-image, including:

[0182] selecting, from the first target region of interest number series corresponding to the first diagonal point, the first target region of interest number series corresponding to the second diagonal point, the second target region of interest number series corresponding to the first diagonal point, and the second target region of interest number series corresponding to the second diagonal point, a number series with the least number of contained regions of interest as the reference number series;

[0183] sequentially taking the regions of interest contained in the reference number series as the object, judging whether the other target region of interest number series includes the overlapping regions of interest, until the overlapping regions of interest in the first target region of interest number series and the second target region of interest number series are found, and the overlapping regions of interest are determined as the target region of interest in the target sub-image.

[0184] Here, the other target region of interest number series can be the other three number series except the reference number series from the first target region of interest number series corresponding to the first diagonal point, the first target region of interest number series corresponding to the second diagonal point, the second target region of interest number series corresponding to the first diagonal point, and the second target region of interest number series corresponding to the second diagonal point.

[0185] In the embodiments of the present application, the number series with the least number of contained regions of interest is selected from the first target region of interest number series corresponding to the first diagonal point, the first target region of interest number series corresponding to the second diagonal point, the second target region of interest number series corresponding to the first diagonal point, and the second target region of interest number series corresponding to the second diagonal point as the reference number series, and the regions of interest contained in the reference sequence are sequentially traversed in the other target region of interest number series to find the overlapping regions of interest in the first target region of interest number series corresponding to the first diagonal point, the first target region of interest number series corresponding to the second diagonal point, the second target region of interest number series corresponding to the first diagonal point, and the second target region of interest number series corresponding to the second diagonal point, and the overlapping regions of interest are taken as the target region of interest in the target sub-image.

[0186] In the above embodiment, the number series with the least number of regions of interest is taken as the reference number series, and other number series are traversed to find the target region of interest. In this way, the target region of interest can be found with the least amount of calculation, the search time is shortened, the search efficiency is improved, and the defect detection time is shortened.

[0187] In some embodiments, the regions of interest included in the reference number series are sequentially taken as objects, and it is determined whether the other target region of interest number series includes coincident regions of interest. This can be implemented as follows:

[0188] The first target region of interest number series corresponding to the first diagonal point, the first target region of interest number series corresponding to the second diagonal point, the second target region of interest number series corresponding to the first diagonal point, and the second target region of interest number series corresponding to the second diagonal point are sorted according to the number of regions of interest;

[0189] A region of interest included in the number series corresponding to the sequence number 1 after sorting is traversed in the number series corresponding to the sequence number 2 to find whether the number series corresponding to the sequence number 2 includes a coincident region of interest;

[0190] In the case where the number series corresponding to the sequence number 2 includes a coincident region of interest, the number series corresponding to the sequence number 3 is traversed to find whether the number series corresponding to the sequence number 3 includes a coincident region of interest;

[0191] In the case where the number series corresponding to the sequence number 3 includes a coincident region of interest, the number series corresponding to the sequence number 4 is traversed to find whether the number series corresponding to the sequence number 4 includes a coincident region of interest;

[0192] In the case where the number series corresponding to the sequence number 2 does not include a coincident region of interest, the number series corresponding to the sequence number 3 does not include a coincident region of interest, or the number series corresponding to the sequence number 4 is traversed, the next region of interest included in the number series corresponding to the sequence number 1 after sorting is returned to traverse the other target region of interest number series until all the regions of interest included in the number series corresponding to the sequence number 1 are traversed.

[0193] Here, the number series corresponding to the sequence number 1 can be the number series with the least number of regions of interest among the first target region of interest number series corresponding to the first diagonal point, the first target region of interest number series corresponding to the second diagonal point, the second target region of interest number series corresponding to the first diagonal point, and the second target region of interest number series corresponding to the second diagonal point. The number series corresponding to the sequence number 2 can be the number series with the second least number of regions of interest among the four number series, the number series corresponding to the sequence number 3 can be the number series with the second most number of regions of interest among the four number series, and the number series corresponding to the sequence number 4 can be the number series with the most number of regions of interest among the four number series.

[0194] In the embodiments of the present application, the number of the interested regions in the number sequence with the least number of interested regions is traversed to find whether there is an overlapping interested region in the number sequence with the second least number of interested regions. If an overlapping interested region is found, the number sequence with the third least number of interested regions is traversed to find whether there is an overlapping interested region. If an overlapping interested region is found, the number sequence with the most number of interested regions is traversed to find whether there is an overlapping interested region. If there is no overlapping interested region in any other number sequence or the number sequence with the most number of interested regions is traversed, the following step of traversing other number sequences is returned.

[0195] In the above embodiments, the sorting according to the number of regions is performed, and then the sorting order is traversed, which is beneficial to reduce the calculation amount.

[0196] In some embodiments, the sequence elements in the first target interested region sequence and the second target interested region sequence can include interested region numbers, so that whether the other target interested region sequence includes an overlapping interested region can be determined by comparing the numbers of the interested regions in the sequences.

[0197] In some embodiments, the level of the defect detection parameter corresponding to each interested region is determined according to the number of suspected defect pixels in each target interested region and the number of features included in each target interested region, including:

[0198] The level of the defect detection parameter corresponding to each interested region is determined according to the ratio of the number of suspected defect pixels to the number of features in each target interested region.

[0199] Here, the number of features can refer to the number of image features included in the target interested region. The more the number of features, the more complex the image. The less the number of features, the flatter the image. The ratio of the number of suspected defect pixels to the number of features and the level of the defect detection parameter can be positively correlated. That is, when the ratio is larger, the level of the defect detection parameter is higher, and thus the defect detection parameter is more sensitive. When the ratio is smaller, the level of the defect detection parameter is lower, and thus the defect detection parameter is less sensitive.

[0200] In the embodiments of the present application, the ratio of the number of suspected defect pixels to the number of features in each target interested region can be calculated, and the level of the defect detection parameter corresponding thereto can be determined according to the correspondence between the ratio and the level of the defect detection parameter.

[0201] Here, the correspondence between the ratio and the defect detection parameter level can be set according to the defect detection requirement. For example, when the ratio is not greater than a first threshold value, the defect detection parameter level is the insensitive level; when the ratio is greater than the first threshold value and less than a second threshold value, the defect detection parameter level is the general level; when the ratio is not less than a third threshold value and less than a fourth threshold value, the defect detection parameter level is the sensitive level; and when the ratio is not less than the fourth threshold value and less than a fifth threshold value, the defect detection parameter level is the super-sensitive level. The threshold values can be set according to the defect detection requirement.

[0202] In the above embodiment, the ratio of the number of suspected defect pixels to the number of features in each target region of interest is used to determine the defect detection parameter level corresponding to each region of interest. In this way, the relationship between each feature of the target region of interest and the defect detection parameter level can be further quantified, and the accuracy of the defect detection parameter level can be improved.

[0203] In some embodiments, before S21, the defect detection method further includes:

[0204] The edge part images of the target object are acquired by rotating the target object along the edge of the target object, and the edge part images are used as the target object images respectively.

[0205] Here, rotating the target object along the edge of the target object can be understood as rotating at least one round around the edge of the target object. Each edge part image can include corresponding imaging of a partial edge part of the target object. Using the edge part images as the target object images can be understood as one edge part image can be one target object image.

[0206] In the embodiments of the present application, the image acquisition device or the defect detection device divides the edge part of the target object into multiple segments, rotates the target object along the edge of the target object, and acquires images of each edge part segment to obtain multiple edge part images, and each edge part image is used as a target object image respectively.

[0207] In the above embodiment, the edge part of the target object is divided into multiple segments to acquire multiple edge part images, and each edge part image is used as a target object image for detection. In this way, the local edge part of the target object can be enlarged, the fineness of the edge part image can be improved, the calculation amount of a single image can be reduced, and the detection efficiency and accuracy of the die surface defect can be further improved.

[0208] In some embodiments, after S23, according to the number of suspected defect pixels in each target region of interest and the number of features contained in each target region of interest, the defect detection parameter level corresponding to each region of interest is determined, the detection method can further include:

[0209] The target regions of interest are detected based on the defect detection parameter levels corresponding to the target regions of interest, to obtain real defects of the target regions of interest.

[0210] Here, the defect detection parameters of the target regions of interest can be set according to the defect detection parameter levels corresponding to the target regions of interest, and whether the suspected defects in the target regions of interest are real defects can be determined according to the defect detection parameters, to obtain real defects of the target regions of interest.

[0211] In the above embodiment, the defect detection parameters of the target regions of interest are determined according to the defect detection parameter levels corresponding to the target regions of interest, and the suspected defects in the target regions of interest are detected according to the respective defect detection parameters, to obtain real defects of the target regions of interest. In this way, the defect detection parameters are adapted to the image features, and the defect detection accuracy is improved.

[0212] In order to have a more overall understanding of the defect detection method provided in the present application, another defect detection method can be applied to a defect detection device as shown in Figure 1 In the embodiments of the present application, the target object image can be a partial or global image of a die, and the target reference points are the top-left corner point and the bottom-right corner point of the region of interest. Figure 6 A flowchart of another defect detection method provided in the embodiments of the present application is shown in FIG. 6. As shown in FIG. 6, the defect detection method includes the following steps. Figure 6

[0213] S61, obtaining a target object image.

[0214] S62, dividing a region of interest according to image features of the target object image, and obtaining first position information of each region of interest in the target object image.

[0215] Here, the first position information can include one or more of the following: a number of the region of interest, corner point coordinate information of the region of interest, and length and width of the region of interest.

[0216] S63, dividing the target object image into a plurality of sub-images, and obtaining second position information of each sub-image.

[0217] Here, the target object image can be divided into M*N sub-images, and M and N are both positive integers greater than 1.

[0218] S64, associating each region of interest with each sub-image based on the first position information of each region of interest and the second position information of each sub-image, to obtain an association relationship between the region of interest and the sub-image.

[0219] ​S65, acquire third position information of each suspected defect pixel in the target object image, and determine the target sub-image to which each suspected defect pixel belongs according to the third position information and the second position information.

[0220] S66, determine the target region of interest to which each suspected defect pixel in each target sub-image belongs based on the association relationship.

[0221] S67, determine the defect detection parameter level corresponding to each target region of interest according to the number of suspected defect pixels in each target region of interest and the number of features contained in each target region of interest.

[0222] In the above embodiment, the target object image is divided into a plurality of small size sub-images, according to the position of the suspected defect pixel, it is determined which sub-image it belongs to, and according to the associated region of interest of the sub-image, it is determined the target region of interest to which the suspected defect pixel belongs, and according to the characteristics of the target region of interest, the defect detection parameter level corresponding to it is determined, in this way, the defect detection parameter level is set adaptively, which is beneficial to subsequent defect detection according to the characteristics of the target region of interest, and improves the detection accuracy. And it can reduce the calculation amount in the parameter setting process, shorten the defect detection time, and improve the defect detection efficiency.

[0223] In some embodiments, S66, based on the association relationship, respectively determine the target region of interest to which each suspected defect pixel in each target sub-image belongs, which can include:

[0224] For each target sub-image, the following steps are performed: based on the association relationship, determine the region of interest associated with the target sub-image;

[0225] The regions of interest associated with the target sub-image are respectively arranged in ascending order according to the size of the X-axis coordinate of the upper left corner point, the Y-axis coordinate of the upper left corner point, the X-axis coordinate of the lower right corner point, and the Y-axis coordinate of the lower right corner point, to form a first array (i.e. the first array corresponding to the upper left corner point), a second array (i.e. the second array corresponding to the upper left corner point), a third array (i.e. the first array corresponding to the lower right corner point), and a fourth array (i.e. the second array corresponding to the lower right corner point);

[0226] Based on the third position information of the suspected defect pixel, find the first reference array element from the first array, find the second reference array element from the second array, find the third reference array element from the third array, and find the fourth reference array element from the fourth array;

[0227] The coincident region of interest in the first array, the second array, the third array and the fourth array is taken as the target region of interest to which the suspected defect pixel belongs.

[0228] Here, the upper left corner point X-axis coordinate of the region of interest associated with the target sub-image is sorted from small to large, and stored in a first array. The stored information can include one or more of the region of interest number, the upper left corner coordinate, the length and width of the region of interest. The upper left corner point Y-axis coordinate of the region of interest associated with the target sub-image is sorted from small to large, and stored in a second array. The stored information can include one or more of the region of interest number, the upper left corner coordinate, the length and width of the region of interest. The right lower corner point X-axis coordinate of the region of interest associated with the target sub-image is sorted from small to large, and stored in a third array. The stored information can include one or more of the region of interest number, the right lower corner coordinate, the length and width of the region of interest. The right lower corner point Y-axis coordinate of the region of interest associated with the target sub-image is sorted from small to large, and stored in a fourth array. The stored information can include one or more of the region of interest number, the right corner coordinate, the length and width of the region of interest.

[0229] The first reference array element is the Ath array element in the first array. In the first array, the upper left corner point X-axis coordinate of the Ath array element is not greater than the X-axis coordinate of the suspected defective pixel, and the upper left corner point X-axis coordinate of the A+1th array element is greater than the X-axis coordinate of the suspected defective pixel. The second reference array element is the Bth array element in the second array. In the second array, the upper left corner point Y-axis coordinate of the Bth array element is not greater than the Y-axis coordinate of the suspected defective pixel, and the upper left corner point Y-axis coordinate of the B+1th array element is greater than the Y-axis coordinate of the suspected defective pixel.

[0230] The third reference array element is the Cth array element in the third array. In the third array, the right lower corner point X-axis coordinate of the C-1th array element is not greater than the X-axis coordinate of the suspected defective pixel, and the right lower corner point X-axis coordinate of the Cth array element is greater than the X-axis coordinate of the suspected defective pixel. The fourth reference array element is the Dth array element in the fourth array. In the fourth array, the right lower corner point Y-axis coordinate of the D-1th array element is not greater than the Y-axis coordinate of the suspected defective pixel, and the right lower corner point Y-axis coordinate of the Dth array element is greater than the Y-axis coordinate of the suspected defective pixel.

[0231] The first array element to the Ath array element in the first array are taken as a first sequence. The first array element to the Bth array element in the second array are taken as a second sequence. The Cth array element to the last array element in the third array are taken as a third sequence. The Dth array element to the last array element in the fourth array are taken as a third sequence.

[0232] In the embodiments of the present application, the first reference array element, the second reference array element, the third reference array element and the fourth reference array element can be found by dichotomy. In addition, the method for finding the overlapping region of interest is similar to the method for finding the overlapping region of interest in the foregoing embodiments, which will not be described here.

[0233] In yet another aspect of the embodiments of the present application, a computer program product is provided, which includes computer instructions, and when the computer instructions are executed by a processor, the computer instructions implement the defect detection method according to any of the embodiments of the present application, and the same technical effects can be achieved by the defect detection method according to the foregoing embodiments.

[0234] In yet another aspect of the embodiments of the present application, a defect detection device is provided. Figure 7 A structural schematic diagram of a defect detection device according to an embodiment of the present application is shown. As shown in the figure, the defect detection device includes a processor 71 and a memory 72, and the memory 72 stores a computer program executable by the processor 71, and when the computer program is executed by the processor 71, the computer program implements the defect detection method according to any of the embodiments of the present application, and the same technical effects can be achieved by the defect detection method according to the foregoing embodiments. To avoid repetition, details are not described herein. Figure 7

[0235] The embodiments of the present application further provide a computer readable storage medium, and the computer readable storage medium stores a computer program. The computer program is executed by a processor to implement each process of the defect detection method embodiments and achieve the same technical effects. To avoid repetition, details are not described herein. The computer readable storage medium includes, for example, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, etc.

[0236] The above merely describes the specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, and all the changes or replacements should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.​

Claims

1. A defect detection method, characterized in that, include: Obtain the first location information of the region of interest in the target object image; Obtain the third position information of the suspected defective pixel in the target object image, and determine the target sub-image to which the suspected defective pixel belongs based on the third position information and the second position information of the multiple sub-images formed by dividing the target object image; Based on the first location information of the region of interest associated with the target sub-image, the target reference points within the region of interest are sorted according to their X-axis coordinates to form a first array, and then sorted according to their Y-axis coordinates to form a second array. Based on the third location information of the suspected defective pixel contained in the target sub-image, determine the first target region of interest sequence in the first array that is adjacent to the suspected defective pixel, and the second target region of interest sequence in the second array that is adjacent to the suspected defective pixel; The regions of interest that overlap in the first and second target regions of interest sequences are determined as the target regions of interest in the target sub-image; the target regions of interest are the regions of interest to which the suspected defective pixels belong. Based on the number of suspected defective pixels in each target region of interest and the number of features contained in each target region of interest, the defect detection parameter level corresponding to each target region of interest is determined.

2. The defect detection method according to claim 1, characterized in that, Before obtaining the third location information of suspected defective pixels in the target object image, and determining the target sub-image to which the suspected defective pixel belongs based on the third location information and the second location information of multiple sub-images formed by dividing the target object image, the method further includes: The target object image is divided into multiple sub-images, and the second position information of the multiple sub-images is obtained; Based on the first location information of the region of interest and the second location information of the plurality of sub-images, each region of interest is associated with the plurality of sub-images to obtain the association relationship between the region of interest and the sub-images.

3. The defect detection method according to claim 2, characterized in that, Before sorting the target reference points within the region of interest according to the first location information associated with the target sub-image to form a first array, and sorting them according to the Y-axis coordinates of the target reference points to form a second array, the method further includes: For each target sub-image, a region of interest associated with the target sub-image is determined based on the association relationship.

4. The defect detection method as described in claim 1, characterized in that, The first array is formed by sorting the target reference points within the region of interest based on the first location information of the region of interest associated with the target sub-image, according to the X-axis coordinates of the target reference points within the region of interest, and the second array is formed by sorting the target reference points based on the Y-axis coordinates of the target reference points, including: The target reference points include the first corner points of the region of interest. Based on the first position information of the region of interest associated with the target sub-image, the points are sorted according to the X-axis coordinates of the first corner points to form a first array, and then sorted according to the Y-axis coordinates of the first corner points to form a second array; or... The target reference points include a first corner point and a second corner point of the region of interest, wherein the second corner point is the diagonal point of the first corner point. Based on the first position information of the region of interest associated with the target sub-image, the first corner points are sorted according to their X-axis coordinates to form a first array, and the second corner points are sorted according to their Y-axis coordinates to form a second array. The second corner points are sorted according to their X-axis coordinates to form a first array, and the second corner points are sorted according to their Y-axis coordinates to form a second array.

5. The defect detection method according to claim 4, characterized in that, The step of sorting the first corner points of the region of interest associated with the target sub-image based on their first location information, forming a first array corresponding to the first corner points, and sorting them based on their Y-axis coordinates, forming a second array corresponding to the first corner points, includes: Based on the first position information of the region of interest associated with the target sub-image, and based on the X-axis coordinate of the first corner point of the region of interest, the X-axis coordinates are sorted in ascending or descending order. One or more of the number of each region of interest, the coordinate value of the first corner point, and the length and width of the region of interest are used as the corresponding array elements to form the first array corresponding to the first corner point. Based on the Y-axis coordinate of the first corner point of the region of interest, the Y-axis coordinate is sorted in ascending or descending order, and one or more of the number of each region of interest, the coordinate value of the first corner point, and the length and width of the region of interest are used as the corresponding array elements to form a second array corresponding to the first corner point. The process involves sorting the second corner points based on their X-axis coordinates within the region of interest to form a first array, and sorting them based on their Y-axis coordinates within the region of interest to form a second array, including: Based on the X-axis coordinate of the second corner point of the region of interest, the X-axis coordinates are sorted in ascending or descending order. One or more of the following are used as array elements: the number of each region of interest, the coordinate value of the second corner point, and the length and width of the region of interest, to form the first array corresponding to the second corner point. Based on the Y-axis coordinate of the second corner point of the region of interest, the Y-axis coordinates are sorted in ascending or descending order, and one or more of the following are used as array elements: the number of each region of interest, the coordinate value of the second corner point, and the length and width of the region of interest, to form a second array corresponding to the second corner point.

6. The defect detection method according to claim 1, characterized in that, The step of determining a first target region of interest (ROI) sequence in the first array adjacent to the suspected defective pixel and a second target ROI sequence in the second array adjacent to the suspected defective pixel based on the third location information of the suspected defective pixel contained in the target sub-image includes: Based on the third location information of the suspected defective pixels contained in the target sub-image; Based on the size of the X coordinate of the suspected defective pixel and the X-axis coordinate of the middle array element of the first array, the array element in front of or behind the first array is taken as the comparison object to determine the first target region of interest sequence in the first array that is adjacent to the suspected defective pixel. Based on the Y-coordinate of the suspected defective pixel and the X-coordinate of the middle array element of the second array, the array elements in front of or behind the second array are taken as comparison objects to determine the second target region of interest sequence in the second array that is adjacent to the suspected defective pixel.

7. The defect detection method as described in claim 1, characterized in that, The step of determining the overlapping regions of interest in the first and second target regions of interest sequences as the target regions of interest in the target sub-image includes: Select the sequence containing the fewest regions of interest from the first target region of interest sequence and the second target region of interest sequence as the reference sequence; Using the regions of interest contained in the reference sequence as objects, determine whether other target regions of interest sequences include overlapping regions of interest, until overlapping regions of interest are found in the first target region of interest sequence and the second target region of interest sequence, and determine them as target regions of interest in the target sub-image.

8. The defect detection method according to any one of claims 1 to 7, characterized in that, The step of determining the defect detection parameter level corresponding to each target region of interest based on the number of suspected defect pixels and the number of features contained in each target region of interest includes: The defect detection parameter level corresponding to each target region of interest is determined based on the ratio of the number of suspected defective pixels to the number of features within each target region of interest.

9. The defect detection method according to any one of claims 1 to 7, characterized in that, Before obtaining the first location information of the region of interest in the target object image, the method further includes: The edge images of the target object are acquired sequentially by rotating along the edge of the target object, and each edge image is used as the target object image.

10. A computer program product, characterized in that, When the instructions in the computer program product are executed by the processor of the defect detection device, the defect detection device performs the defect detection method as described in any one of claims 1 to 9.

11. A defect detection device, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program that can be executed by the processor, and the computer program, when executed by the processor, implements the defect detection method as described in any one of claims 1 to 9.

12. A detection system, characterized in that, Includes image acquisition equipment and the defect detection equipment as described in claim 11. The image acquisition device is used to rotate and scan around the edge of the target object, and sequentially acquire images of the edge parts of the target object as the target object image and send them to the defect detection device.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by the controller, implements the defect detection method as described in any one of claims 1 to 9.

Images