Chip positioning method, device, equipment and medium based on multi-feature fusion

Abstract

The present application relates to a chip positioning method, device, computer equipment and storage medium based on multi-feature fusion. On the basis of edge detection and conventional positioning, two target straight line segments with the largest and smallest endpoint coordinate values ​​are selected from the set of straight line segments in four directions (left, right, up and down), which are the two straight line segments closest to the two vertices in this direction, so as to most accurately represent the four vertices of the chip image, and the intersection points of the eight straight line segments are further calculated according to the obtained eight target straight line segments as the four vertices of the chip, and finally the shape and position of the chip in the image are accurately determined.

Description

Technical Field

[0001] The present application relates to the field of machine vision technology, and in particular to a chip positioning method, device, equipment and medium based on multi-feature fusion. Background Art

[0002] Chip usually refers to a semiconductor product made of silicon wafer materials. With the development of technology, chips are widely used in computers, digital terminals and other digital appliances, and are an important part of the structure of modern information society. The chip manufacturing process is complex, and chip defect detection is an important part of chip production and manufacturing. Chip positioning is the first step in chip detection, and its positioning accuracy directly affects the accuracy and reliability of subsequent chip defect detection. Therefore, accurate positioning of chip images plays a vital role.

[0003] There are two types of chip positioning methods, namely, deep learning methods based on semantic segmentation and classical methods based on edge detection. With the development of deep learning technology, researchers have proposed a series of chip positioning algorithms based on deep learning target recognition (such as YOLO model) and semantic segmentation (such as UNET model). Both methods are very computationally intensive and usually require graphics card support, which greatly increases the cost of the positioning system. In addition, users are required to collect and train chip samples in advance, which makes it difficult to meet the requirements of flexible detection in industrial production. At the same time, although the target recognition method based on deep learning can identify the position of the chip, its detection result is to mark the chip area in the form of a horizontal rectangular frame. When the chip is tilted at an angle, it cannot completely fit the chip outline and cannot meet the requirements of precise positioning. When flexible detection is performed in industrial production, the classical method based on edge detection is usually used.

[0004] However, the above-mentioned traditional classical method based on edge detection has the problem of insufficient chip positioning accuracy. Summary of the invention

[0005] Based on this, it is necessary to provide a chip positioning method based on multi-feature fusion, a chip positioning device based on multi-feature fusion, a computer device and a computer-readable storage medium to address the above technical problems.

[0006] In order to achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0007] On the one hand, the present invention provides a chip positioning method based on multi-feature fusion, comprising:

[0008] The chip image of the chip to be positioned is obtained and the geometric dimension data of the chip template is obtained; the geometric dimension data of the chip template includes the lengths and angles of each side of the chip.

[0009] The straight line segment i of the chip image is extracted using edge detection; i=1,2,…,N.

[0010] According to the position of the straight line segment i relative to the center point of the chip image and the direction of the straight line segment i on the chip image, the straight line segment i is divided into the first straight line segment set, the second straight line segment set, the third straight line segment set or the fourth straight line segment set.

[0011] Preliminary selection boxes of the chip image are determined according to the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set.

[0012] Each preliminary selection box is screened according to the geometric size data to obtain each alternative selection box.

[0013] Each selected box is obtained by filtering from each candidate box through edge color detection.

[0014] Split each selection box into straight line segments j; j=1,2,…,M, M<N.

[0015] According to the position of the straight line segment j relative to the center point of the chip image and the direction of the straight line segment j on the chip image, the straight line segment j is divided into a left vertical line set, a right vertical line set, an upper horizontal line set or a lower horizontal line set.

[0016] Eight target straight line segments are obtained by selecting the straight line segment j with the largest endpoint coordinate value and the smallest endpoint coordinate value from the left vertical line set, the right vertical line set, the upper horizontal line set and the lower horizontal line set respectively.

[0017] The upper left vertex, upper right vertex, lower left vertex and lower right vertex of the chip image are calculated according to the eight target straight line segments, and the precise positioning frame of the chip image is determined.

[0018] On the other hand, a chip positioning device based on multi-feature fusion is also provided, including a chip defect detection host computer, the chip defect detection host computer includes:

[0019] A collection module is used to obtain a chip image of the chip to be positioned and obtain geometric dimension data of the chip template; the geometric dimension data of the chip template includes the lengths and angles of each side of the chip;

[0020] An extraction module, used to extract a straight line segment i of a chip image by using edge detection; i=1, 2, ..., N;

[0021] A first division module, used for dividing the straight line segment i into a first straight line segment set, a second straight line segment set, a third straight line segment set or a fourth straight line segment set according to the position of the straight line segment i relative to the center point of the chip image and the direction of the straight line segment i on the chip image;

[0022] A combination module, used for determining each preliminary selection frame of the chip image according to the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set;

[0023] The first screening module is used to screen each preliminary selection box according to the geometric size data to obtain each candidate selection box;

[0024] The second screening module is used to screen out each selected frame from each candidate frame by edge color detection;

[0025] A splitting module is used to split each selected box into straight line segments j; j = 1, 2, ..., M, M < N;

[0026] A second division module is used to divide the straight line segment j into a left vertical line set, a right vertical line set, an upper horizontal line set or a lower horizontal line set according to the position of the straight line segment j relative to the center point of the chip image and the direction of the straight line segment j on the chip image;

[0027] A selection module is used to select the straight line segment j with the largest endpoint coordinate value and the smallest endpoint coordinate value from the left vertical line set, the right vertical line set, the upper horizontal line set and the lower horizontal line set to obtain eight target straight line segments;

[0028] The positioning module is used to calculate the upper left vertex, upper right vertex, lower left vertex and lower right vertex of the chip image according to the eight target straight line segments, and determine the precise positioning frame of the chip image.

[0029] On the other hand, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned chip positioning method based on multi-feature fusion when executing the computer program.

[0030] On the other hand, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned chip positioning method based on multi-feature fusion are implemented.

[0031] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0032] The above chip positioning method, device, equipment and medium based on multi-feature fusion, based on edge detection and conventional positioning, respectively selects two target straight line segments with the largest and smallest endpoint coordinate values ​​from the set of straight line segments in four directions (left, right, up and down), which are the two straight line segments closest to the two vertices in this direction, so as to most accurately represent the positions of the four vertices of the chip image, and further calculates the intersection points of the eight straight line segments according to the obtained eight target straight line segments, as the four vertices of the chip, and finally accurately determines the shape and position of the chip in the image. This method not only utilizes edge detection and conventional positioning methods, adds constraints on image color information near the edge, thereby filtering out erroneous candidate frames, and improves environmental adaptability and anti-interference capabilities, but also considers the set of straight line segments corresponding to the four directions (left, right, up and down), fully mines the mutual position information of the line segments in the image, selects the best straight line segments representing the vertices, and then uses these best straight line segments to recalculate the four vertices to obtain the possible optimal vertex selection, and finally realizes accurate positioning of the chip. In the present invention, it is necessary not only to base on the previous positioning results (the selected boxes obtained), but also to break and reorganize the external information of these selected boxes (split each selected box into straight line segments j), and regenerate the positioning closest to the vertex (precise positioning box), so as to achieve sub-pixel positioning accuracy. This method belongs to a new positioning idea. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the conventional technology, the drawings required for use in the embodiments or the conventional technology descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0034] Figure 1 Schematic diagram of a process of a chip positioning method based on multi-feature fusion in one embodiment;

[0035] Figure 2 A schematic diagram of a process for determining each preliminary selection box step in an embodiment;

[0036] Figure 3 Schematic diagram of the module structure of a chip positioning device based on multi-feature fusion in one embodiment. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0038] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application.

[0039] It should be noted that the reference to "embodiment" in this document means that the specific features, structures or characteristics described in conjunction with the embodiment may be included in at least one embodiment of the present invention. The presentation of the phrase at various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments.

[0040] In the process of studying and implementing this application, the inventor pointed out that chip defect detection usually requires locating the position of the target chip first, and the positioning accuracy of the chip has a great influence on the result of defect detection. Inaccurate positioning can easily lead to false detection, thereby reducing the accuracy of defect detection. The inventor found that the previous chip positioning method determines the position of the chip through edge detection and template matching. This method is suitable for positioning detection in the ideal state where the chip is not damaged in the packaging factory. However, after the user manufacturer receives the chips shipped by the packaging factory, various tests and inspections must be performed one by one to ensure the quality of the received chips. At this time, the chip may be worn on its edge due to repeated plugging and unplugging of the test socket of the packaging factory machine or collision during transportation, etc., resulting in a small amount of dents or burrs. The previous methods did not take this situation into account, resulting in the algorithm being able to detect multiple edge lines on the same outer edge of the chip, thereby forming multiple selection frames of a chip. Therefore, it is necessary to further analyze and screen the selection frames obtained by the traditional algorithm to eliminate the chip vertex positions with large errors. In addition, the actual production environment is very complex and changeable. For example, the color, material, and lighting of the trays used to detect chips from different batches may not be consistent, which may cause the edges and lines generated during edge detection and line detection to change, potentially leading to incorrect candidate positions.

[0041] Based on this, the present invention provides a chip positioning method, device, equipment and medium based on multi-feature fusion. On the basis of edge detection and conventional positioning, two target straight line segments with the largest and smallest endpoint coordinate values ​​are selected from the set of straight line segments in four directions (left, right, up and down), respectively. These are the two straight line segments closest to the two vertices in this direction, so as to most accurately represent the positions of the four vertices of the chip image. The intersection points of the eight straight line segments are further calculated according to the obtained eight target straight line segments, as the four vertices of the chip, and finally the shape and position of the chip in the image are accurately determined. This method not only utilizes edge detection and conventional positioning methods, adds constraints on image color information near the edge, thereby filtering out erroneous candidate frames, and improves environmental adaptability and anti-interference capabilities, but also considers the set of straight line segments corresponding to the four directions (left, right, up and down), fully explores the mutual position information of the line segments in the image, selects the best straight line segments representing the vertices, and then uses these best straight line segments to recalculate the four vertices to obtain the possible optimal vertex selection, and finally realizes the accurate positioning of the chip. In the present invention, it is necessary not only to base on the previous positioning results (the selected boxes obtained), but also to break and reorganize the external information of these selected boxes (split each selected box into straight line segments j), and regenerate the positioning closest to the vertex (precise positioning box), so as to achieve sub-pixel positioning accuracy. This method belongs to a new positioning idea.

[0042] The following will describe the implementation of the present invention in detail with reference to the accompanying drawings in the embodiment diagram of the present invention.

[0043] In one embodiment, Figure 1 As shown, the embodiment of the present application provides a chip positioning method based on multi-feature fusion, including the following processing steps S11-S20:

[0044] S11, obtaining a chip image of the chip to be positioned and obtaining geometric dimension data of the chip template; the geometric dimension data of the chip template includes the lengths and angles of each side of the chip.

[0045] It is understandable that the chip image can be obtained by taking a photo of the chip to be positioned in the tray. Due to the limitation of the size and position of the tray, the chip image will be located in the middle of the photo, but not necessarily in the middle. The four sides are not necessarily horizontal or vertical, but may be tilted, but the inclination angle will not be greater than 45 degrees. Before positioning, it is also necessary to obtain the geometric dimension data of the chip template. This can be done by taking a photo of the chip template and calculating the length and angle of the quadrilateral, extracting the geometric dimension data of the chip template from the database, or directly measuring and recording the size of a physical chip and manually inputting it into the system. The geometric dimension data of the chip template includes the data of the length and angle of each side of the chip.

[0046] S12, extracting straight line segment i of the chip image by edge detection; i=1,2,…,N.

[0047] It can be understood that before edge detection is performed on the acquired chip image, the color chip image can be converted into a grayscale image, the purpose of which is to remove color information and retain only brightness information in preparation for edge detection. The Canny algorithm can be applied to the obtained grayscale image to detect all edge contours in the photo, and then the straight line detection algorithm can be further used to locate and extract N straight line segments that make up the edge contour, where N is a positive integer.

[0048] S13, dividing the straight line segment i into a first straight line segment set, a second straight line segment set, a third straight line segment set or a fourth straight line segment set according to the position of the straight line segment i relative to the center point of the chip image and the direction of the straight line segment i on the chip image.

[0049] It can be understood that the position of the straight line segment i relative to the center point of the chip image: for example, the straight line segment i is located on the left, right, above or below the center point of the image. The direction of the straight line segment i on the chip image: for example, the straight line segment i is horizontal, vertical or inclined. Considering the above two factors, the vertical lines located on the left side of the center point of the image are divided into the first straight line segment set, the vertical lines located on the right side of the center point of the image are divided into the second straight line segment set, the horizontal lines located above the center point of the image are divided into the third straight line segment set, and the horizontal lines located below the center point of the image are divided into the fourth straight line segment set.

[0050] S14, determining preliminary selection boxes of the chip image according to the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set.

[0051] It can be understood that a straight line segment is selected from each of the first straight line segment set, the second straight line segment set, the third straight line segment set, and the fourth straight line segment set, and the intersection of these four straight line segments is calculated to form a quadrilateral as a preliminary selection frame. Due to wear and other reasons, the outer edge of the chip may be dented or damaged, resulting in multiple intermittent straight line segments being detected on one outer edge. These straight line segments are all part of the outer edge of the chip, and the inclination angles of these straight line segments may be slightly different, and they may not necessarily be connected into a straight line, resulting in the preliminary selection frames formed by these edge intersections not overlapping. In addition, there are also trays, gaskets and other debris in the chip photo, and other straight line segments will also be detected. Permutations and combinations can be used to determine all quadrilaterals that can be formed in the four straight line segment sets without duplication or omission, thereby obtaining the preliminary selection frames of the chip image.

[0052] S15, screening each preliminary selection box according to the geometric size data to obtain each candidate selection box.

[0053] It can be understood that the side length and angle of each preliminary selection box are calculated, compared with the geometric size data of the chip template, and the preliminary selection boxes close to the geometric size data of the template are retained, thereby obtaining various candidate boxes.

[0054] S16, selecting each selected frame from each candidate frame by edge color detection.

[0055] It can be understood that during edge detection and line detection, some invalid edges and lines will be detected in external environments such as trays. These invalid edge lines are mixed with the edge lines of the chip, which may produce some invalid candidate boxes. To this end, it is necessary to check the color constraints of the corresponding positions of the edges of each candidate box in the chip image to eliminate the candidate boxes with mismatched colors. For example, for each edge of each candidate box, extract the color near the center of the candidate box. If the inner color matches the template feature and the outer color matches the tray feature, it means that the edge is valid; due to the influence of foreign matter and edge damage, the colors of the four edges of the candidate box may not be exactly the same, so the average color of the edge segment is compared with the template color. If the difference between the average color of the four edges of the candidate box and the template color is within the tolerable threshold range, the candidate box is considered valid, and each valid candidate box is retained to obtain each selected box. The following step is to obtain a precise positioning box from multiple selected boxes.

[0056] S17, split each selected frame into straight line segments j; j=1, 2, ..., M, M<N.

[0057] It can be understood that by disassembling the quadrilateral selection box, M straight line segments are obtained, where M is a positive integer. Because each preliminary selection box is screened twice to obtain each selection box, M<N.

[0058] S18, dividing the straight line segment j into a left vertical line set, a right vertical line set, an upper horizontal line set or a lower horizontal line set according to the position of the straight line segment j relative to the center point of the chip image and the direction of the straight line segment j on the chip image.

[0059] It can be understood that the vertical lines located on the left side of the image center point are divided into the left vertical line set, the vertical lines located on the right side of the image center point are divided into the right vertical line set, the horizontal lines located above the image center point are divided into the upper horizontal line set, and the horizontal lines located below the image center point are divided into the horizontal line set.

[0060] S19, respectively selecting the straight line segment j with the largest endpoint coordinate value and the smallest endpoint coordinate value from the left vertical line set, the right vertical line set, the upper horizontal line set and the lower horizontal line set to obtain eight target straight line segments.

[0061] It can be understood that the straight line segment with the largest ordinate is selected from the left vertical line set, which can be recorded as L 1MAX, select the straight line segment with the smallest ordinate from the left vertical line set, which can be recorded as L 1MIN , select the straight line segment with the largest ordinate from the right vertical line set, which can be recorded as L 2MAX , select the straight line segment with the smallest ordinate from the right vertical line set, which can be recorded as L 2MIN , select the straight line segment with the largest horizontal coordinate from the upper horizontal line set, which can be recorded as L 3MAX , select the straight line segment with the smallest horizontal coordinate from the upper horizontal line set, which can be recorded as L 3MIN , select the straight line segment with the largest horizontal coordinate from the lower horizontal line set, which can be recorded as L 4MAX , select the straight line segment with the smallest horizontal coordinate from the lower horizontal line set, which can be recorded as L 4MIN . Thus, eight target straight line segments are obtained: L 1MAX , L 1MIN , L 2MAX , L 2MIN , L 3MAX , L 3MIN , L 4MAX and L 4MIN , when the straight line segment with the largest ordinate and the straight line segment with the smallest ordinate in the left vertical line set correspond to the same straight line segment, L 1MAX and L 1MIN For the same straight line segment, two of the eight target straight line segments overlap, and other similar situations are understood in the same way.

[0062] S20, calculating the upper left vertex, the upper right vertex, the lower left vertex and the lower right vertex of the chip image according to the eight target straight line segments, and determining the precise positioning frame of the chip image.

[0063] It is understandable that the calculation of L 1MIN and L 3MIN The intersection point is taken as the upper left vertex of the chip image and L is calculated. 2MIN and L 3MAX The intersection point is taken as the upper right vertex of the chip image and L is calculated. 1MAX and L 4MIN The intersection point is taken as the lower left vertex of the chip image and L is calculated. 2MAX and L 4MAX The intersection of the two pixels is taken as the lower right vertex of the chip image, thereby obtaining an accurate positioning frame of the chip image.

[0064] The chip positioning method based on multi-feature fusion, on the basis of edge detection and conventional positioning, selects two target straight line segments with the largest and smallest endpoint coordinate values ​​from the set of straight line segments in four directions (left, right, up, and down), which are the two straight line segments closest to the two vertices in this direction, so as to most accurately represent the positions of the four vertices of the chip image. The intersection of the eight straight line segments is further calculated based on the obtained eight target straight line segments, which are used as the four vertices of the chip, and finally the shape and position of the chip in the image are accurately determined. This method not only utilizes edge detection and conventional positioning methods, adds constraints on image color information near the edge, thereby filtering out erroneous candidate frames, and improves the adaptability to the environment and anti-interference ability, but also considers the set of straight line segments corresponding to the four directions (left, right, up, and down), fully explores the mutual position information of the line segments in the image, selects the best straight line segments representing the vertices, and then uses these best straight line segments to recalculate the four vertices to obtain the possible optimal vertex selection, and finally realizes the accurate positioning of the chip. In the present invention, it is necessary not only to base on the previous positioning results (the selected boxes obtained), but also to break and reorganize the external information of these selected boxes (split each selected box into straight line segments j), and regenerate the positioning closest to the vertex (precise positioning box), so as to achieve sub-pixel positioning accuracy. This method belongs to a new positioning idea.

[0065] In addition, the step of filtering the preliminary selection frame through geometric size data and obtaining the alternative selection frame utilizes the geometric size characteristics of the chip to filter out the frames that do not match the template size from the preliminary selection frame, thereby enhancing the accuracy of positioning and reducing the false detection rate caused by environmental changes. The step of filtering the selected frame through edge color detection introduces edge color features, which can filter out some false edges caused by changes in ambient light, making the positioning of the selected frame more accurate and reliable.

[0066] In one embodiment, Figure 2 As shown, the process of determining each preliminary selection box in the above method of the present invention may include the following processing steps S141-S146:

[0067] S141 , selecting a first straight line segment, a second straight line segment, a third straight line segment and a fourth straight line segment from the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set respectively to form a basic positioning frame.

[0068] It can be understood that the first straight line segment is randomly selected from the first straight line segment set, the second straight line segment is randomly selected from the second straight line segment set, the third straight line segment is randomly selected from the third straight line segment set, and the fourth straight line segment is randomly selected from the fourth straight line segment set, and the quadrilateral formed by the first straight line segment, the second straight line segment, the third straight line segment and the fourth straight line segment is used as the basic positioning frame.

[0069] S142, using other straight line segments in the first straight line segment collection to replace the first straight line segments in the basic positioning frame respectively, to obtain left replacement quadrilateral sets.

[0070] It can be understood that in the first straight line segment set, except for the first straight line segment, one of the remaining straight line segments is selected to replace the first straight line segment in the basic positioning frame to form a left-alternative quadrilateral, and then another remaining straight line segment is selected to replace the first straight line segment in the basic positioning frame to form another left-alternative quadrilateral. This replacement is repeated multiple times until the remaining straight line segments in the first straight line segment set have each replaced the first straight line segment in the basic positioning frame once, forming multiple left-alternative quadrilaterals. These multiple left-alternative quadrilaterals are collected together to form a left-alternative quadrilateral set. During the replacement process, it is ensured that each remaining straight line segment in the first straight line segment set has a replacement opportunity, and finally all possible left-alternative quadrilaterals are obtained, and the left-alternative quadrilateral set is constructed without omission.

[0071] S143, using other straight line segments in the second straight line segment set to replace the second straight line segments in the basic positioning frame respectively, to obtain right replacement quadrilateral sets.

[0072] It can be understood that the process of obtaining the right-alternative quadrilateral sets can refer to the above-mentioned process of obtaining the left-alternative quadrilateral sets, which will not be repeated here.

[0073] S144, using other straight line segments in the third straight line segment set to replace the third straight line segments in the basic positioning frame respectively, to obtain respective replacement quadrilateral sets.

[0074] It can be understood that the process of obtaining each upper substitute quadrilateral set can refer to the process of obtaining the left substitute quadrilateral set mentioned above, which will not be repeated here.

[0075] S145, using other straight line segments in the fourth straight line segment set to replace the fourth straight line segments in the basic positioning frame respectively, to obtain respective replacement quadrilateral sets.

[0076] It can be understood that the process of obtaining the lower substitute quadrilateral sets can refer to the process of obtaining the left substitute quadrilateral set mentioned above, which will not be repeated here.

[0077] S146, using the basic positioning frame, each left substitute quadrilateral set, each right substitute quadrilateral set, each upper substitute quadrilateral set and each lower substitute quadrilateral set as each preliminary selection frame of the chip image.

[0078] It can be understood that the basic positioning frame, the left-alternative quadrilateral sets, the right-alternative quadrilateral sets, the upper-alternative quadrilateral sets and the lower-alternative quadrilateral sets contain all quadrilaterals that can be formed by the straight line segment i, without duplication or omission. These quadrilaterals are used as the preliminary selection frames of the chip image to facilitate the subsequent screening of the preliminary selection frames.

[0079] It should be understood that although Figure 1-2 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1-2 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0080] In one embodiment, Figure 3 As shown, a chip positioning device 200 based on multi-feature fusion is provided, including a chip defect detection host computer 210, and the chip defect detection host computer 210 includes:

[0081] The collection module 211 is used to obtain the chip image of the chip to be positioned and obtain the geometric size data of the chip template; the geometric size data of the chip template includes the lengths and angles of each side of the chip.

[0082] The extraction module 212 is used to extract the straight line segment i of the chip image by edge detection; i=1, 2, ..., N.

[0083] The first division module 213 is used to divide the straight line segment i into the first straight line segment set, the second straight line segment set, the third straight line segment set or the fourth straight line segment set according to the position of the straight line segment i relative to the center point of the chip image and the direction of the straight line segment i on the chip image.

[0084] The combination module 214 is used to determine each preliminary selection box of the chip image according to the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set.

[0085] The first screening module 215 is used to screen each preliminary selection box according to the geometric size data to obtain each candidate selection box.

[0086] The second screening module 216 is used to screen out each selected frame from each candidate frame by edge color detection.

[0087] The splitting module 217 is used to split each selected frame into straight line segments j; j=1, 2, ..., M, M<N.

[0088] The second division module 218 is used to divide the straight line segment j into a left vertical line set, a right vertical line set, an upper horizontal line set or a lower horizontal line set according to the position of the straight line segment j relative to the center point of the chip image and the direction of the straight line segment j on the chip image.

[0089] The selection module 219 is used to select the straight line segment j with the largest endpoint coordinate value and the smallest endpoint coordinate value from the left vertical line set, the right vertical line set, the upper horizontal line set and the lower horizontal line set to obtain eight target straight line segments.

[0090] The positioning module 220 is used to calculate the upper left vertex, the upper right vertex, the lower left vertex and the lower right vertex of the chip image according to the eight target straight line segments, and determine the precise positioning frame of the chip image.

[0091] The chip positioning device based on multi-feature fusion, on the basis of edge detection and conventional positioning, selects two target straight line segments with the largest and smallest endpoint coordinate values ​​from the set of straight line segments in four directions (left, right, up, and down), respectively. These are the two straight line segments closest to the two vertices in this direction, thereby most accurately representing the positions of the four vertices of the chip image. The intersection of these eight straight line segments is further calculated based on the obtained eight target straight line segments, as the four vertices of the chip, and finally accurately determines the shape and position of the chip in the image. This method not only utilizes edge detection and conventional positioning methods, adds constraints on image color information near the edge, thereby filtering out erroneous candidate frames, and improves environmental adaptability and anti-interference capabilities, but also considers the set of straight line segments corresponding to the four directions (left, right, up, and down), fully exploits the mutual position information of the line segments in the image, selects the best straight line segments representing the vertices, and then uses these best straight line segments to recalculate the four vertices to obtain the possible optimal vertex selection, and finally achieves accurate positioning of the chip. In the present invention, it is necessary to not only base on the previous positioning results (obtained selected frames), but also to break and reorganize the external information of these selected frames (split each selected frame into straight line segments j), and regenerate the positioning closest to the vertex (precise positioning frame), so as to achieve sub-pixel positioning accuracy. This method belongs to a brand-new positioning idea. In addition, the step of selecting the preliminary selection frame through geometric size data and obtaining the candidate frame utilizes the geometric size characteristics of the chip, and filters out the frames that do not match the template size from the preliminary selection frame, thereby enhancing the accuracy of positioning and reducing the false detection rate caused by environmental changes; the step of selecting the selected frame through edge color detection introduces edge color features, which can filter out some false edges caused by changes in ambient light, making the positioning of the selected frame more accurate and reliable.

[0092] In one embodiment, the chip positioning device 200 based on multi-feature fusion, the combination module 214 includes:

[0093] The basic combination submodule is used to select the first straight line segment, the second straight line segment, the third straight line segment and the fourth straight line segment from the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set to form a basic positioning frame. The left replacement submodule is used to use other straight line segments in the first straight line segment set to replace the first straight line segment in the basic positioning frame to obtain left replacement quadrilateral sets. The right replacement submodule is used to use other straight line segments in the second straight line segment set to replace the second straight line segment in the basic positioning frame to obtain right replacement quadrilateral sets. The upper replacement submodule is used to use other straight line segments in the third straight line segment set to replace the third straight line segment in the basic positioning frame to obtain upper replacement quadrilateral sets. The lower replacement submodule is used to use other straight line segments in the fourth straight line segment set to replace the fourth straight line segment in the basic positioning frame to obtain lower replacement quadrilateral sets. The final combination submodule is used to use the basic positioning frame, left replacement quadrilateral sets, right replacement quadrilateral sets, upper replacement quadrilateral sets and lower replacement quadrilateral sets as preliminary selection frames of chip images.

[0094] In one embodiment, the above-mentioned chip positioning device 200 based on multi-feature fusion also includes a parallel light source, a camera, a machine and a tray. The parallel light source is used to illuminate the chip to be positioned, the camera is used to photograph the chip to be positioned and output the chip image, the machine is used to carry the tray, and the tray is used to load the chip to be positioned.

[0095] It can be understood that the parallel light source is used to illuminate the chip to be positioned with parallel light. The parallel light can reduce the effects of scattering or diffraction of light when irradiating the target, which is conducive to taking clear images. The camera is used to photograph the chip to be positioned under the illumination of the parallel light source and output the corresponding chip image. The machine provides a stable platform to support the tray. The machine can be a mechanical device such as an optical workbench. The tray is placed on the machine to load multiple chips to be positioned, which can easily position different chips.

[0096] For the specific definition of the chip positioning device based on multi-feature fusion, please refer to the definition of the chip positioning method based on multi-feature fusion mentioned above, which will not be repeated here. Each module in the above-mentioned chip positioning device based on multi-feature fusion can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0097] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following processing steps are implemented:

[0098] Acquire a chip image of the chip to be positioned and acquire geometric dimension data of the chip template; the geometric dimension data of the chip template includes the lengths and angles of each side of the chip; use edge detection to extract a straight line segment i of the chip image; i=1,2,…,N; divide the straight line segment i into a first straight line segment set, a second straight line segment set, a third straight line segment set or a fourth straight line segment set according to the position of the straight line segment i relative to the center point of the chip image and the direction of the straight line segment i on the chip image; determine each preliminary selection box of the chip image according to the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set; screen each preliminary selection box according to the geometric dimension data to obtain each candidate box; Color detection selects each selected frame from each candidate frame; splits each selected frame into a straight line segment j; j=1,2,…,M,M<N; divides the straight line segment j into a left vertical line set, a right vertical line set, an upper horizontal line set or a lower horizontal line set according to the position of the straight line segment j relative to the center point of the chip image and the direction of the straight line segment j on the chip image; selects the straight line segment j with the largest endpoint coordinate value and the smallest endpoint coordinate value from the left vertical line set, the right vertical line set, the upper horizontal line set and the lower horizontal line set to obtain eight target straight line segments; calculates the upper left vertex, the upper right vertex, the lower left vertex and the lower right vertex of the chip image according to the eight target straight line segments, and determines the precise positioning frame of the chip image.

[0099] The above computer device, based on edge detection and conventional positioning, selects two target straight line segments with the largest and smallest endpoint coordinate values ​​from the set of straight line segments in four directions (left, right, up, and down), which are the two straight line segments closest to the two vertices in this direction, so as to most accurately represent the positions of the four vertices of the chip image, and further calculates the intersection points of the eight straight line segments according to the obtained eight target straight line segments, as the four vertices of the chip, and finally accurately determines the shape and position of the chip in the image. This method not only utilizes edge detection and conventional positioning methods, adds constraints on image color information near the edge, thereby filtering out erroneous candidate frames, and improves environmental adaptability and anti-interference capabilities, but also considers the set of straight line segments corresponding to the four directions (left, right, up, and down), fully explores the mutual position information of the line segments in the image, selects the best straight line segments representing the vertices, and then uses these best straight line segments to recalculate the four vertices to obtain the possible optimal vertex selection, and finally achieves accurate positioning of the chip. In the present invention, it is necessary to not only base on the previous positioning results (obtained selected frames), but also to break and reorganize the external information of these selected frames (split each selected frame into straight line segments j), and regenerate the positioning closest to the vertex (precise positioning frame), so as to achieve sub-pixel positioning accuracy. This method belongs to a brand-new positioning idea. In addition, the step of selecting the preliminary selection frame through geometric size data and obtaining the candidate frame utilizes the geometric size characteristics of the chip, and filters out the frames that do not match the template size from the preliminary selection frame, thereby enhancing the accuracy of positioning and reducing the false detection rate caused by environmental changes; the step of selecting the selected frame through edge color detection introduces edge color features, which can filter out some false edges caused by changes in ambient light, making the positioning of the selected frame more accurate and reliable.

[0100] In one embodiment, the above-mentioned computer device, when the processor executes the computer program, further implements the following processing steps:

[0101] Select the first straight line segment, the second straight line segment, the third straight line segment and the fourth straight line segment from the first straight line segment set, the second straight line segment set, the third straight line segment and the fourth straight line segment set to form a basic positioning frame. Use the other straight line segments in the first straight line segment set to replace the first straight line segment in the basic positioning frame to obtain the left substitute quadrilateral sets. Use the other straight line segments in the second straight line segment set to replace the second straight line segment in the basic positioning frame to obtain the right substitute quadrilateral sets. Use the other straight line segments in the third straight line segment set to replace the third straight line segment in the basic positioning frame to obtain the upper substitute quadrilateral sets. Use the other straight line segments in the fourth straight line segment set to replace the fourth straight line segment in the basic positioning frame to obtain the lower substitute quadrilateral sets. Use the basic positioning frame, the left substitute quadrilateral sets, the right substitute quadrilateral sets, the upper substitute quadrilateral sets and the lower substitute quadrilateral sets as the primary selection frames of the chip image.

[0102] In one embodiment, the above-mentioned computer device is a chip defect detection host computer.

[0103] For the specific limitations of the above-mentioned computer device, please refer to the limitations of the chip positioning method based on multi-feature fusion in the above text, which will not be repeated here.

[0104] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following processing steps are implemented:

[0105] Acquire a chip image of the chip to be positioned and acquire geometric dimension data of the chip template; the geometric dimension data of the chip template includes the lengths and angles of each side of the chip; use edge detection to extract a straight line segment i of the chip image; i=1,2,…,N; divide the straight line segment i into a first straight line segment set, a second straight line segment set, a third straight line segment set or a fourth straight line segment set according to the position of the straight line segment i relative to the center point of the chip image and the direction of the straight line segment i on the chip image; determine each preliminary selection box of the chip image according to the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set; screen each preliminary selection box according to the geometric dimension data to obtain each candidate box; Color detection selects each selected frame from each candidate frame; splits each selected frame into a straight line segment j; j=1,2,…,M,M<N; divides the straight line segment j into a left vertical line set, a right vertical line set, an upper horizontal line set or a lower horizontal line set according to the position of the straight line segment j relative to the center point of the chip image and the direction of the straight line segment j on the chip image; selects the straight line segment j with the largest endpoint coordinate value and the smallest endpoint coordinate value from the left vertical line set, the right vertical line set, the upper horizontal line set and the lower horizontal line set to obtain eight target straight line segments; calculates the upper left vertex, the upper right vertex, the lower left vertex and the lower right vertex of the chip image according to the eight target straight line segments, and determines the precise positioning frame of the chip image.

[0106] The computer-readable storage medium, based on edge detection and conventional positioning, selects two target straight line segments with the largest and smallest endpoint coordinate values ​​from the set of straight line segments in four directions (left, right, up, and down), which are the two straight line segments closest to the two vertices in this direction, so as to most accurately represent the positions of the four vertices of the chip image, and further calculates the intersection points of the eight straight line segments according to the obtained eight target straight line segments as the four vertices of the chip, and finally accurately determines the shape and position of the chip in the image. This method not only utilizes edge detection and conventional positioning methods, adds constraints on image color information near the edge, thereby filtering out erroneous candidate frames, and improves environmental adaptability and anti-interference capabilities, but also considers the set of straight line segments corresponding to the four directions (left, right, up, and down), fully exploits the mutual position information of the line segments in the image, selects the best straight line segments representing the vertices, and then uses these best straight line segments to recalculate the four vertices to obtain the possible optimal vertex selection, and finally achieves accurate positioning of the chip. In the present invention, it is necessary to not only base on the previous positioning results (obtained selected frames), but also to break and reorganize the external information of these selected frames (split each selected frame into straight line segments j), and regenerate the positioning closest to the vertex (precise positioning frame), so as to achieve sub-pixel positioning accuracy. This method belongs to a brand-new positioning idea. In addition, the step of selecting the preliminary selection frame through geometric size data and obtaining the candidate frame utilizes the geometric size characteristics of the chip, and filters out the frames that do not match the template size from the preliminary selection frame, thereby enhancing the accuracy of positioning and reducing the false detection rate caused by environmental changes; the step of selecting the selected frame through edge color detection introduces edge color features, which can filter out some false edges caused by changes in ambient light, making the positioning of the selected frame more accurate and reliable.

[0107] In one embodiment, the computer program further implements the following processing steps when executed by a processor:

[0108] Select the first straight line segment, the second straight line segment, the third straight line segment and the fourth straight line segment from the first straight line segment set, the second straight line segment set, the third straight line segment and the fourth straight line segment set to form a basic positioning frame. Use the other straight line segments in the first straight line segment set to replace the first straight line segment in the basic positioning frame to obtain the left substitute quadrilateral sets. Use the other straight line segments in the second straight line segment set to replace the second straight line segment in the basic positioning frame to obtain the right substitute quadrilateral sets. Use the other straight line segments in the third straight line segment set to replace the third straight line segment in the basic positioning frame to obtain the upper substitute quadrilateral sets. Use the other straight line segments in the fourth straight line segment set to replace the fourth straight line segment in the basic positioning frame to obtain the lower substitute quadrilateral sets. Use the basic positioning frame, the left substitute quadrilateral sets, the right substitute quadrilateral sets, the upper substitute quadrilateral sets and the lower substitute quadrilateral sets as the primary selection frames of the chip image.

[0109] For the specific definition of the above-mentioned computer-readable storage medium, please refer to the definition of the chip positioning method based on multi-feature fusion in the above text, which will not be repeated here.

[0110] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0111] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0112] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A chip positioning method based on multi-feature fusion, characterized in that: Includes steps: Acquire a chip image of the chip to be positioned and acquire geometric dimension data of the chip template; the geometric dimension data of the chip template includes the lengths and angles of each side of the chip; Extracting a straight line segment i of the chip image by edge detection; i=1, 2, ..., N; According to the position of the straight line segment i relative to the center point of the chip image and the direction of the straight line segment i on the chip image, dividing the straight line segment i into a first straight line segment set, a second straight line segment set, a third straight line segment set or a fourth straight line segment set; Determine each preliminary selection box of the chip image according to the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set; Screening each of the preliminary selection boxes according to the geometric size data to obtain each of the alternative selection boxes; Filtering each selected frame from each candidate frame by edge color detection; Splitting each of the selected boxes into straight line segments j; j=1,2,…,M,M<N; According to the position of the straight line segment j relative to the center point of the chip image and the direction of the straight line segment j on the chip image, dividing the straight line segment j into a left vertical line set, a right vertical line set, an upper horizontal line set or a lower horizontal line set; Selecting the straight line segment j with the largest endpoint coordinate value and the smallest endpoint coordinate value from the left vertical line set, the right vertical line set, the upper horizontal line set, and the lower horizontal line set to obtain eight target straight line segments; The upper left vertex, the upper right vertex, the lower left vertex and the lower right vertex of the chip image are calculated according to the eight target straight line segments, and the precise positioning frame of the chip image is determined.

2. The chip positioning method based on multi-feature fusion according to claim 1 is characterized in that: The step of determining each preliminary selection frame of the chip image according to the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set comprises: Selecting a first straight line segment, a second straight line segment, a third straight line segment and a fourth straight line segment from the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set respectively to form a basic positioning frame; Use other straight line segments in the first straight line segment collection to replace the first straight line segments in the basic positioning frame respectively, to obtain left-alternative quadrilateral sets; Use other straight line segments in the second straight line segment set to replace the second straight line segments in the basic positioning frame respectively, to obtain right-substitute quadrilateral sets; Use other straight line segments in the third straight line segment set to replace the third straight line segments in the basic positioning frame respectively, to obtain each replacement quadrilateral set; Use other straight line segments in the fourth straight line segment set to replace the fourth straight line segments in the basic positioning frame respectively, to obtain respective replacement quadrilateral sets; The basic positioning frame, the left-alternative quadrilateral sets, the right-alternative quadrilateral sets, the upper-alternative quadrilateral sets and the lower-alternative quadrilateral sets are used as preliminary selection frames of the chip image.

3. A chip positioning device based on multi-feature fusion, characterized in that: Including chip defect detection host computer, the chip defect detection host computer includes: A collection module, used to obtain a chip image of the chip to be positioned and obtain geometric dimension data of the chip template; the geometric dimension data of the chip template includes the lengths and angles of each side of the chip; An extraction module, used for extracting a straight line segment i of the chip image by edge detection; i=1, 2, ..., N; A first division module, configured to divide the straight line segment i into a first straight line segment set, a second straight line segment set, a third straight line segment set or a fourth straight line segment set according to the position of the straight line segment i relative to the center point of the chip image and the direction of the straight line segment i on the chip image; A combination module, used for determining each preliminary selection box of the chip image according to the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set; A first screening module, used for screening each of the preliminary selection boxes according to the geometric size data to obtain each candidate selection box; A second screening module is used to screen out each selected frame from each candidate frame by edge color detection; A splitting module, used for splitting each selected frame into straight line segments j; j = 1, 2, ..., M, M < N; A second division module is used to divide the straight line segment j into a left vertical line set, a right vertical line set, an upper horizontal line set or a lower horizontal line set according to the position of the straight line segment j relative to the center point of the chip image and the direction of the straight line segment j on the chip image; A selection module, used for selecting the straight line segment j with the largest endpoint coordinate value and the smallest endpoint coordinate value from the left vertical line set, the right vertical line set, the upper horizontal line set and the lower horizontal line set to obtain eight target straight line segments; A positioning module is used to calculate the upper left vertex, the upper right vertex, the lower left vertex and the lower right vertex of the chip image according to the eight target straight line segments, and determine the precise positioning frame of the chip image.

4. The chip positioning device based on multi-feature fusion according to claim 3 is characterized in that: The combined module comprises: A basic combination submodule, used for selecting a first straight line segment, a second straight line segment, a third straight line segment and a fourth straight line segment from the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set respectively to form a basic positioning frame; A left replacement submodule, used to replace the first straight line segments in the basic positioning frame with other straight line segments in the first straight line segment collection, to obtain left replacement quadrilateral sets; a right replacement submodule, used to replace the second straight line segments in the basic positioning frame with other straight line segments in the second straight line segment set, to obtain right replacement quadrilateral sets; An upper replacement submodule, used to use other straight line segments in the third straight line segment set to replace the third straight line segments in the basic positioning frame respectively, to obtain upper replacement quadrilateral sets; A lower replacement submodule, used to use other straight line segments in the fourth straight line segment set to replace the fourth straight line segments in the basic positioning frame respectively, to obtain respective lower replacement quadrilateral sets; The final combination submodule is used to use the basic positioning frame, the left substitute quadrilateral sets, the right substitute quadrilateral sets, the upper substitute quadrilateral sets and the lower substitute quadrilateral sets as the primary selection frames of the chip image.

5. The chip positioning device based on multi-feature fusion according to claim 3, characterized in that: It also includes a parallel light source, a camera, a machine and a tray, wherein the parallel light source is used to illuminate the chip to be positioned, the camera is used to photograph the chip to be positioned and output the chip image, the machine is used to carry the tray, and the tray is used to load the chip to be positioned.

6. A computer device comprising a memory and a processor, characterized in that: The memory stores a computer program, and when the processor executes the computer program, the following processing steps are implemented: Acquire a chip image of the chip to be positioned and acquire geometric dimension data of the chip template; the geometric dimension data of the chip template includes the lengths and angles of each side of the chip; Extracting a straight line segment i of the chip image by edge detection; i=1, 2, ..., N; According to the position of the straight line segment i relative to the center point of the chip image and the direction of the straight line segment i on the chip image, dividing the straight line segment i into a first straight line segment set, a second straight line segment set, a third straight line segment set or a fourth straight line segment set; Determine each preliminary selection box of the chip image according to the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set; Screening each of the preliminary selection boxes according to the geometric size data to obtain each of the alternative selection boxes; Filtering each selected frame from each candidate frame by edge color detection; Splitting each of the selected boxes into straight line segments j; j=1,2,…,M,M<N; According to the position of the straight line segment j relative to the center point of the chip image and the direction of the straight line segment j on the chip image, dividing the straight line segment j into a left vertical line set, a right vertical line set, an upper horizontal line set or a lower horizontal line set; Selecting the straight line segment j with the largest endpoint coordinate value and the smallest endpoint coordinate value from the left vertical line set, the right vertical line set, the upper horizontal line set, and the lower horizontal line set to obtain eight target straight line segments; The upper left vertex, the upper right vertex, the lower left vertex and the lower right vertex of the chip image are calculated according to the eight target straight line segments, and the precise positioning frame of the chip image is determined.

7. The computer device according to claim 6, characterized in that When the processor executes the computer program, the following processing steps are also implemented: Selecting a first straight line segment, a second straight line segment, a third straight line segment and a fourth straight line segment from the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set respectively to form a basic positioning frame; Use other straight line segments in the first straight line segment collection to replace the first straight line segments in the basic positioning frame respectively, to obtain left-alternative quadrilateral sets; Use other straight line segments in the second straight line segment set to replace the second straight line segments in the basic positioning frame respectively, to obtain right-substitute quadrilateral sets; Use other straight line segments in the third straight line segment set to replace the third straight line segments in the basic positioning frame respectively, to obtain each replacement quadrilateral set; Use other straight line segments in the fourth straight line segment set to replace the fourth straight line segments in the basic positioning frame respectively, to obtain respective replacement quadrilateral sets; The basic positioning frame, the left-alternative quadrilateral sets, the right-alternative quadrilateral sets, the upper-alternative quadrilateral sets and the lower-alternative quadrilateral sets are used as preliminary selection frames of the chip image.

8. The computer device according to claim 6, characterized in that The computer device is a chip defect detection host computer.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the following processing steps are implemented: Acquire a chip image of the chip to be positioned and acquire geometric dimension data of the chip template; the geometric dimension data of the chip template includes the lengths and angles of each side of the chip; Extracting a straight line segment i of the chip image by edge detection; i=1, 2, ..., N; According to the position of the straight line segment i relative to the center point of the chip image and the direction of the straight line segment i on the chip image, dividing the straight line segment i into a first straight line segment set, a second straight line segment set, a third straight line segment set or a fourth straight line segment set; Determine each preliminary selection box of the chip image according to the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set; Screening each of the preliminary selection boxes according to the geometric size data to obtain each of the alternative selection boxes; Filtering each selected frame from each candidate frame by edge color detection; Splitting each of the selected boxes into straight line segments j; j=1,2,…,M,M<N; According to the position of the straight line segment j relative to the center point of the chip image and the direction of the straight line segment j on the chip image, dividing the straight line segment j into a left vertical line set, a right vertical line set, an upper horizontal line set or a lower horizontal line set; Selecting the straight line segment j with the largest endpoint coordinate value and the smallest endpoint coordinate value from the left vertical line set, the right vertical line set, the upper horizontal line set, and the lower horizontal line set to obtain eight target straight line segments; The upper left vertex, the upper right vertex, the lower left vertex and the lower right vertex of the chip image are calculated according to the eight target straight line segments, and the precise positioning frame of the chip image is determined.

10. The computer-readable storage medium according to claim 9, wherein: When the computer program is executed by the processor, the following processing steps are also implemented: Selecting a first straight line segment, a second straight line segment, a third straight line segment and a fourth straight line segment from the first straight line segment set, the second straight line segment set, the third straight line segment set and the fourth straight line segment set respectively to form a basic positioning frame; Use other straight line segments in the first straight line segment collection to replace the first straight line segments in the basic positioning frame respectively, to obtain left-alternative quadrilateral sets; Use other straight line segments in the second straight line segment set to replace the second straight line segments in the basic positioning frame respectively, to obtain right-substitute quadrilateral sets; Use other straight line segments in the third straight line segment set to replace the third straight line segments in the basic positioning frame respectively, to obtain each replacement quadrilateral set; Use other straight line segments in the fourth straight line segment set to replace the fourth straight line segments in the basic positioning frame respectively, to obtain respective replacement quadrilateral sets; The basic positioning frame, the left-alternative quadrilateral sets, the right-alternative quadrilateral sets, the upper-alternative quadrilateral sets and the lower-alternative quadrilateral sets are used as preliminary selection frames of the chip image.

Images