Method and system for detecting WAT abnormal needle marks
Patent Information
- Application Number
- CN202510397236.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-31
- Publication Date
- 2025-07-18
Smart Images
Figure CN120339209A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of semiconductor manufacturing, and in particular to a method and system for detecting abnormal pin marks in WAT. Background Art
[0002] When performing electrical parameter tests on wafers, due to probe card wear or needle alignment problems on the machine, the probe card may be misaligned, inserted too deep or too shallow, which will ultimately result in the inability to receive data in the WAT (Wafer Acceptance Test) and the failure of acceptance and shipment.
[0003] Currently, after problems are found in the WAT test data, usually engineers need to manually collect pin mark images, analyze and troubleshoot the abnormal causes based on the pin marks (such as pin mark deviation, pin mark being too large, pin mark being too shallow), and then make corresponding adjustments. The entire process has poor feedback timeliness, low efficiency and serious lag, and cannot discover problems and make improvements in a timely manner, easily causing test losses.
[0004] Based on this, how to improve the analysis efficiency of abnormal pin marks in WAT tests, ensure the test quality of WAT, and avoid test losses caused by feedback lag has become a technical problem that needs to be solved by those skilled in the art. Summary of the Invention
[0005] The purpose of the present invention is to provide a method and system for detecting abnormal pin marks in WAT, so as to solve the problems in the prior art that the detection process takes a long time, has low efficiency, cannot discover problems and make improvements in a timely manner, and easily causes test losses.
[0006] To achieve the above purpose, the present invention provides a method for detecting abnormal pin marks in WAT, including:
[0007] Collecting a pinning image;
[0008] Obtaining the closed contour data of the test pads on the pinning image;
[0009] Based on the closed contour data, positioning all the test pads on the pinning image;
[0010] Performing rectangular detection on the test pads, extracting the size data of the test pads, and comparing the ratio of the pin marks on the test pads with a preset range to determine whether the pin marks are normal and classifying the abnormal pin marks.
[0011] Optionally, the performing rectangular detection on the test pads, extracting the size data of the test pads, and comparing the ratio of the pin marks on the test pads with a preset range to determine whether the pin marks are normal and classifying the abnormal pin marks includes:
[0012] When a complete rectangle cannot be detected on the test pad, or there are no pin marks on the test pad, it is determined that the pin marks on the current test pad are abnormal pin marks;
[0013] When a complete rectangle is detected on the test pad and there are pin marks on the test pad, the size data of the test pad is extracted, and the pin marks are segmented and extracted to compare the ratio of the pin marks on the test pad with a preset range.
[0014] Optionally, the extracting the size data of the test pad and segmenting and extracting the pin marks to compare the ratio of the pin marks on the test pad with a preset range includes:
[0015] Extract the width, height, and area data of the test pad and the pin marks respectively;
[0016] Compare the ratio of the width of the pin mark to the width of the test pad with a first preset range, the ratio of the height of the pin mark to the height of the test pad with a second preset range, and the ratio of the area of the pin mark to the area of the test pad with a third preset range respectively;
[0017] When the ratios of the width, height, and area of the pin mark on the test pad are respectively within the corresponding preset ranges, it is determined that the pin mark is a normal pin mark;
[0018] When any one of the ratios of the width, height, and area of the pin mark on the test pad is outside the corresponding preset range, it is determined that the pin mark is an abnormal pin mark.
[0019] Optionally, when a complete rectangle cannot be detected on the test pad, or there are no pin marks on the test pad, it is determined that the type of the abnormal pin mark is pin mark offset;
[0020] When any one of the ratios of the width, height, and area of the pin mark on the test pad is greater than the maximum value of the corresponding preset range, it is determined that the type of the abnormal pin mark is pin mark too large;
[0021] When any one of the ratios of the width, height, and area of the pin mark on the test pad is less than the minimum value of the corresponding preset range, it is determined that the type of the abnormal pin mark is pin mark too shallow.
[0022] Optionally, the first preset range includes: 0.05 < W / W b <0.3; where W is the width of the pin mark and W b is the width of the test pad;
[0023] The second preset range includes: 0.05 < H / H b< 0.8; where H is the height of the needle mark, H b is the height of the test pad;
[0024] The third preset range includes: 0.05 < S / S b < 0.3; where S is the height of the needle mark, S b is the area of the test pad.
[0025] Optionally, obtaining the closed contour data of the test pads on the needle insertion image includes:
[0026] Performing edge detection, contour dilation, and contour detection on the needle insertion image in sequence through a digital image processing algorithm to obtain the closed contour data of the test pads on the needle insertion image.
[0027] Optionally, positioning all the test pads on the needle insertion image based on the closed contour data includes:
[0028] Cleaning the test pad coordinates from the closed contour data;
[0029] Based on the test pad coordinates, obtaining the width, height, and area of the test pad, and forming an aspect ratio range and an area range;
[0030] Positioning all the test pads on the needle insertion image based on the aspect ratio range and the area range.
[0031] Optionally, the aspect ratio range includes: 4 / 3 ≤ H b / W b ≤ 3 / 2; where H b is the height of the test pad, and W b is the width of the test pad;
[0032] The area range includes: 480 ≤ S b ≤ 520, where S b is the area of the test pad.
[0033] Optionally, extracting the width and height of the needle mark or the test pad includes:
[0034] Respectively extracting the maximum value and the minimum value in the horizontal direction, and the maximum value and the minimum value in the vertical direction in the set of contour points;
[0035] Taking the difference between the maximum value and the minimum value in the horizontal direction as the width; taking the difference between the maximum value and the minimum value in the vertical direction as the height.
[0036] To achieve the above object, the present invention further provides a system for detecting abnormal needle marks in WAT, which is applied to the method for detecting abnormal needle marks in WAT as described above, and includes:
[0037] An acquisition unit for acquiring the needle insertion image;
[0038] A preprocessing unit, communicatively connected to the acquisition unit, for obtaining the closed contour data of the test pads on the needle insertion image;
[0039] A positioning unit, communicatively connected to the preprocessing unit, for positioning all the test pads on the needle insertion image based on the closed contour data;
[0040] A processing unit, communicatively connected to the positioning unit, for performing rectangular detection on the test pads, extracting the size data of the test pads, and comparing the ratio of the needle marks on the test pads with a preset range to determine whether the needle marks are normal and classify the abnormal needle marks.
[0041] Compared with the existing methods for detecting abnormal needle marks, the method and system for detecting abnormal needle marks in WAT provided by the present application have the following advantages:
[0042] The method for detecting abnormal needle marks in WAT provided by the present application can automatically acquire the needle insertion image. After the WAT test is completed, it can locate all the test pads through an image processing algorithm, and combine rectangular detection and compare the ratio of the needle marks on the test pads with a preset range, thereby judging the needle mark situation, realizing the function of automatically identifying abnormal needle marks, and being able to classify different abnormal needle marks. It no longer requires engineers to manually acquire images and analyze, improving the work efficiency of engineers, optimizing the abnormal feedback process of WAT tests, ensuring the quality of WAT tests, and avoiding test losses caused by delayed feedback; at the same time, through the closed contour data, the test pads can be accurately located, excluding the interference of the background area on abnormal detection and improving the accuracy of abnormal detection. Description of the Drawings
[0043] Figure 1 It is a flowchart of the method for detecting abnormal needle marks in WAT provided by an embodiment of the present invention;
[0044] Figure 2a It is a schematic diagram of the contour of the test pads on the first original needle insertion image provided by an embodiment of the present invention;
[0045] Figure 2b It is a schematic diagram of the contour of the test pads on the second original needle insertion image provided by an embodiment of the present invention;
[0046] Figure 2cSchematic diagram of the contour of the test pad on the needling image after contour expansion provided by the embodiment of the present invention;
[0047] Figure 2d Schematic diagram for obtaining the necessary and contour data of the test pad provided by the embodiment of the present invention;
[0048] Figure 3 Schematic diagram for extracting the maximum and minimum values provided by the embodiment of the present invention;
[0049] Figure 4 Schematic diagram for performing rectangular detection on the test pad provided by the embodiment of the present invention;
[0050] Figure 5 Schematic diagram of the abnormal detection result provided by the embodiment of the present invention;
[0051] Figure 6 Schematic diagram of the system for detecting abnormal needle marks of WAT provided by the embodiment of the present invention.
[0052] In the figure, the descriptions of the respective reference numerals are as follows:
[0053] 1 - Test pad; 2 - Needle mark;
[0054] 10 - Acquisition unit; 20 - Preprocessing unit; 30 - Positioning unit; 40 - Processing unit. Detailed implementation manners
[0055] To make the objectives, advantages and features of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the accompanying drawings are all in very simplified forms and are not drawn to scale, and are only used to conveniently and clearly assist in explaining the objectives of the embodiments of the present invention. In addition, the structures shown in the accompanying drawings are often part of the actual structures. In particular, the accompanying drawings need to show different emphases and sometimes different scales are used.
[0056] As used in this specification, the singular forms "a", "an" and "the" include plural referents, the term "or" is generally used in the sense of including "and / or", the term "several" is generally used in the sense of including "at least one", the term "at least two" is generally used in the sense of including "two or more", in addition, the terms "first", "second", "third" are used for descriptive purposes only and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first", "second", "third" may explicitly or implicitly include one or at least two of such features, "one end" and "the other end" and "proximal end" and "distal end" generally refer to two corresponding parts, which include not only the endpoints, the terms "mounted", "connected", "coupled" shall be understood in a broad sense, for example, it may be a fixed connection, a detachable connection, or integral; it may be directly connected, or indirectly connected through an intermediate medium, it may be the internal communication of two components or the interaction relationship between two components. In addition, as used in this specification, one component being disposed on another component generally only means that there is a connection, coupling, cooperation or transmission relationship between the two components, and the two components may be directly or indirectly connected, coupled, cooperated or transmitted through an intermediate component, and cannot be construed as indicating or implying the spatial position relationship between the two components, that is, one component may be inside, outside, above, below or on one side of the other component, etc. in any orientation, unless otherwise explicitly specified in the content. The terms "upper", "lower", "top", "bottom" are generally relative position relationships arranged in the direction of gravity; the terms "vertical, vertical direction" generally refer to the direction along the direction of gravity, which is generally perpendicular to the ground, the "horizontal, horizontal plane direction" is generally along the direction parallel to the ground; for those of ordinary skill in the art, the specific meanings of the above terms in this specification can be understood according to specific circumstances.
[0057] The object of the present invention is to provide a method and system for detecting abnormal needle marks of WAT, so as to solve the problems in the prior art that the detection process takes a long time, has low efficiency, cannot find problems and make improvements in time, and is likely to cause test losses.
[0058] Those skilled in the art can understand that in the semiconductor manufacturing process, the WAT test is a key electrical test link before the wafer leaves the factory, used to verify whether the process parameters of the chip meet the design specifications. And the determination of pin marks is particularly important in this process because the pin marks left when the probe card contacts the wafer may affect the subsequent packaging process or chip performance. When the metal probes on the probe card contact the wafer pads, physical marks may be left due to mechanical pressure and current loading. Among them, normal pin marks are slightly indented or discolored and do not affect the electrical properties; while abnormal pin marks may cause too deep scratches, pad damage, metal layer peeling or probe contamination, thus affecting the accuracy of packaging and test results. The existing detection of abnormal pin marks is that after finding problems in the test data, engineers manually collect pin mark images and analyze and determine abnormal pin marks, which is inefficient and the feedback timeliness is seriously lagged, easily causing test losses. Based on this, this embodiment provides a method and system for detecting WAT abnormal pin marks, which can automatically identify abnormal pin marks after the WAT test is completed, classify different abnormal pin marks, improve the efficiency of abnormal detection, and at the same time optimize the abnormal feedback process to avoid test losses caused by feedback lag.
[0059] Please refer to Figure 1 , the present invention provides a method for detecting WAT abnormal pin marks, including:
[0060] Step S1: Collect pin insertion images;
[0061] Step S2: Obtain the closed contour data of test pad 1 on the pin insertion image;
[0062] Step S3: Based on the closed contour data, locate all test pads 1 on the pin insertion image;
[0063] Step S4: Perform rectangular detection on test pad 1, extract the size data of test pad 1, and compare the ratio of pin mark 2 on test pad 1 with the preset range to determine whether the pin mark 2 is normal, and classify the abnormal pin mark 2.
[0064] As an optional embodiment, in step S1, an imaging device (such as a camera or a video camera, etc.) on the test machine table can be used to collect pin insertion images in real time during the test process. Among them, in this embodiment, the pin insertion image refers to the image of test pad 1 containing pin mark 2 after the probe is inserted into test pad 1 and the test is completed.
[0065] Further, in step S2, a digital image processing algorithm can be used to process the needle insertion image obtained in step S1, and the closed contour data of the test pad 1 in the needle insertion image can be obtained. Optionally, the process of using the digital image processing algorithm to process the needle insertion image may include: binary processing, Gaussian noise reduction, edge detection, contour dilation, and contour detection operations. Those skilled in the art can understand that Gaussian noise reduction can eliminate noise interference while retaining the overall contour and edge features of the image, making the image smoother; binary processing can convert a color image into an image with only black and white colors, reducing the data complexity, and at the same time, it can highlight the key areas through threshold segmentation.
[0066] In this embodiment, after binary processing and Gaussian noise reduction, edge detection is performed on the obtained contour image of the test pad 1, and the results are as Figure 2a and Figure 2b shown. Multiple breakpoints appear on the contour of the test pad 1 (as shown in the black frames in Figure 2a and Figure 2b ), and these breakpoints will affect the subsequent acquisition of the width and height of the test pad 1, and further affect the accuracy of anomaly detection. Therefore, it is also necessary to perform contour dilation on the contour image of the test pad 1 until, as shown in Figure 2c , the dilated contour lines are connected (as shown in the black frame in Figure 2c ), so that the contour is closed. During this process, the engineer can set the dilation ratio based on the actual situation as long as the dilated contour can be closed. Finally, contour detection operations are performed on the dilated and closed contour image, as shown in Figure 2d , to obtain the closed contour data of the test pad 1. It should be noted that to reduce the influence of the background pattern during the processing, a partial image can be intercepted from the needle insertion image as the processing object before processing.
[0067] Please refer to Figure 3 , in an alternative embodiment, in step S3, first clean the closed contour data obtained in step S2 to obtain the coordinates of the test pad 1; based on the coordinates of the test pad 1, obtain the width, height, and area of the test pad 1, and form an aspect ratio range and an area range; based on the aspect ratio range and the area range, locate all the test pads 1 on the needle insertion image. It should be noted that the above coordinates of the test pad 1 can be set in advance before the test or can be automatically generated.
[0068] Optionally, after cleaning to obtain the coordinates of the test pad 1, the maximum and minimum values in the horizontal and vertical directions can be taken from the contour point set; the difference between the maximum and minimum values in the horizontal direction is used as the width; the difference between the maximum and minimum values in the vertical direction is used as the height. Take Figure 3For example, define the horizontal direction as the X-axis and the vertical direction as the Y-axis. In the set of contour points, take the maximum value max(x) and the minimum value min(x) in the horizontal direction. The width H of the above-mentioned test pad 1 b = max(x) - min(x); take the maximum value max(y) and the minimum value min(y) in the vertical direction. The height W of the above-mentioned test pad 1 b = max(y) - min(y); correspondingly, the area S of the above-mentioned test pad 1 b = H b ×W b .
[0069] In this embodiment, the aspect ratio range is 4 / 3 ≤ H b / W b ≤ 3 / 2; the area range is 480 ≤ S b ≤ 520. During the positioning process, all the test pads 1 on the zapping needle image can be accurately located according to the aspect ratio, area, and the spacing between adjacent test pads 1. With such a configuration, the test pads 1 can be accurately located, the interference of the background area on the anomaly detection can be removed, and the accuracy of the anomaly detection is improved.
[0070] In step S4, please refer to Figure 4 , it is necessary to detect the images of all the test pads 1 located in step S3 one by one to distinguish the normal needle marks 2 from the abnormal needle marks 2 and classify the abnormal needle marks 2 for subsequent adjustment of the WAT test parameters and the corresponding probes.
[0071] In some embodiments of the present application, it is necessary to perform a rectangular detection for each test pad 1. When the zapping needle image of the test pad 1 is as shown in Figure 4 (Ⅰ) below, the needle mark 2 is not zapped at the center of the test pad 1 but is offset, resulting in the inability to detect a complete rectangle at the edge of the test pad 1 during the rectangular detection. At this time, it should be determined that the needle mark 2 on the current test pad 1 is an abnormal needle mark 2 and classified as a needle mark 2 offset; if the above situation is not handled and adjusted in time, it may occur that the needle mark 2 is completely offset to the surrounding devices of the test pad 1. At this time, there may be a situation where there is no needle mark 2 on the test pad 1. Therefore, even if a complete rectangle can be detected on the test pad 1, when there is no needle mark 2 on the test pad 1, it is still determined that the needle mark 2 on the current test pad 1 is an abnormal needle mark 2 and classified as a needle mark 2 offset.
[0072] Optionally, when a complete rectangle is detected on the test pad 1 and there is a needle mark 2 on the test pad 1 (as shown in Figure 4as shown in II in), extract the dimensional data of the test pad 1, and segment and extract the pin marks 2 to compare the ratio of the pin marks 2 on the test pad 1 with the preset range. In this embodiment, the width, height, and area data of the test pad 1 and the pin marks 2 can be extracted separately first; for the width, height, and area data of the test pad 1, reference can be made to the method in step S3, which will not be elaborated in this embodiment. Regarding the width, height, and area data of the pin marks 2, taking Figure 4 as an example, with the horizontal direction as the X-axis and the vertical direction as the Y-axis, after segmenting and extracting the pin marks 2, the maximum value max(x i ) and the minimum value min(x i ) in the horizontal direction can be taken from the contour point set of the pin marks 2. The width W of the above-mentioned pin marks 2 = max(x i ) - min(x i ); take the maximum value max(y i ) and the minimum value min(y i ) in the vertical direction. The height H of the above-mentioned pin marks 2 = max(y i ) - min(y i ); correspondingly, the area S of the above-mentioned pin marks 2 = ∑x i y i .
[0073] Furthermore, compare the relationship between the ratio of the width of the pin marks 2 to the width of the test pad 1 and the first preset range, the relationship between the ratio of the height of the pin marks 2 to the height of the test pad 1 and the second preset range, and the relationship between the ratio of the area of the pin marks 2 to the area of the test pad 1 and the third preset range. In this embodiment, the first preset range is: 0.05 < W / W b < 0.3; the second preset range is: 0.05 < H / H b < 0.8; the third preset range is 0.05 < S / S b < 0.3. In some other embodiments, the first preset range, the second preset range, and the third preset range can also be other reasonable values, which are not limited in this embodiment.
[0074] Further, when the ratios of the width, height, and area of the needle mark 2 on the test pad 1 are respectively within the corresponding preset ranges, the needle mark 2 is determined to be a normal needle mark 2; when any one of the ratios of the width, height, and area of the needle mark 2 on the test pad 1 is outside the corresponding preset range, the needle mark 2 is determined to be an abnormal needle mark 2. Among them, when any one of the ratios of the width, height, and area of the needle mark 2 on the test pad 1 is greater than the maximum value of the corresponding preset range, the type of the abnormal needle mark 2 is determined to be that the needle mark 2 is too large; when any one of the ratios of the width, height, and area of the needle mark 2 on the test pad 1 is less than the minimum value of the corresponding preset range, the type of the abnormal needle mark 2 is determined to be that the needle mark 2 is too shallow. Please refer to Figure 5 , taking test1 to test5 as an example, through the above abnormal detection method, the final results are shown in Table 1 as follows:
[0075] Table 1 Statistics of Abnormal Detection Results
[0076]
[0077] With such a configuration, by automatically collecting the needle-poking images, after the WAT test is completed, all the test pads 1 can be located through the image processing algorithm, and by combining the rectangular detection and comparing the ratio of the needle mark 2 on the test pad 1 with the preset range, the situation of the needle mark 2 can be judged, realizing the function of automatically identifying the abnormal needle mark 2, and being able to classify different abnormal needle marks 2. There is no need for engineers to manually collect images and analyze, improving the work efficiency of engineers, optimizing the abnormal feedback process of the WAT test, ensuring the quality of the WAT test, and avoiding test losses caused by delayed feedback.
[0078] Please refer to Figure 6 , in another embodiment, the present invention further provides a system for detecting abnormal needle marks in WAT, which is applied to the method for detecting abnormal needle marks in WAT as described above, and includes: a collection unit 10 for collecting needle-poking images; a preprocessing unit 20 communicatively connected to the collection unit 10 for obtaining the closed contour data of the test pad 1 on the needle-poking image; a positioning unit 30 communicatively connected to the preprocessing unit 20 for positioning all the test pads 1 on the needle-poking image based on the closed contour data; a processing unit 40 communicatively connected to the positioning unit 30 for performing rectangular detection on the test pad 1, extracting the size data of the test pad 1, and comparing the ratio of the needle mark 2 on the test pad 1 with the preset range to determine whether the needle mark 2 is normal and classify the abnormal needle mark 2.
[0079] It should be noted that in this embodiment, the acquisition unit 10 may be an imaging device on the machine platform, and the preprocessing unit 20, the positioning unit 30, and the processing unit 40 may respectively call digital image processing algorithms to process the needle insertion image. With such a configuration, after the WAT test is completed, the needle insertion image can be directly acquired, the test pad 1 can be located, and the needle mark 2 can be judged, realizing the automatic recognition and classification of abnormal needle marks, optimizing the abnormal feedback process of the WAT test, improving the work efficiency of engineers, ensuring the quality of the WAT test, and avoiding test losses caused by delayed feedback.
[0080] In summary, in the method and system for detecting WAT abnormal needle marks provided in the embodiment of the present invention, the method for detecting WAT abnormal needle marks includes: acquiring a needle insertion image; obtaining the closed contour data of the test pads on the needle insertion image; positioning all the test pads on the needle insertion image based on the closed contour data; performing rectangular detection on the test pads, extracting the size data of the test pads, and comparing the ratio of the needle marks on the test pads with a preset range to determine whether the needle marks are normal and classify the abnormal needle marks.
[0081] With such a configuration, by automatically acquiring the needle insertion image, after the WAT test is completed, all the test pads can be located through the image processing algorithm, and combined with rectangular detection and comparing the ratio of the needle marks on the test pads with the preset range, the needle mark situation can be further judged, realizing the function of automatically identifying abnormal needle marks, and being able to classify different abnormal needle marks. It is no longer necessary for engineers to manually acquire images and analyze, improving the work efficiency of engineers, optimizing the abnormal feedback process of the WAT test, ensuring the quality of the WAT test, and avoiding test losses caused by delayed feedback; at the same time, through the closed contour data, the test pads can be accurately located, excluding the interference of the background area on the abnormal detection, and improving the accuracy of the abnormal detection.
[0082] The above description is only a description of the preferred embodiment of the present invention, and does not limit the scope of the present invention in any way. Any changes and modifications made by those of ordinary skill in the field of the present invention according to the above disclosure are within the protection scope of the claims.
Claims
1. A method for detecting abnormal needle marks on WAT, characterized in that, Including: Collecting acupuncture images; Obtaining the closed contour data of the test pads on the acupuncture image; Based on the closed contour data, locating all the test pads on the acupuncture image; Performing rectangular detection on the test pads, extracting the size data of the test pads, and comparing the ratio of the needle marks on the test pads with a preset range to determine whether the needle marks are normal and classifying the abnormal needle marks.
2. The method for detecting abnormal needle marks of WAT according to claim 1, wherein The performing rectangular detection on the test pads, extracting the size data of the test pads, and comparing the ratio of the needle marks on the test pads with a preset range to determine whether the needle marks are normal and classifying the abnormal needle marks includes: When a complete rectangle cannot be detected on the test pad or there are no needle marks on the test pad, determining that the needle marks on the current test pad are abnormal needle marks; When a complete rectangle is detected on the test pad and there are needle marks on the test pad, extracting the size data of the test pad and performing segmentation extraction on the needle marks to compare the ratio of the needle marks on the test pad with a preset range.
3. The method for detecting abnormal needle marks of WAT according to claim 2, characterized in that, The extracting the size data of the test pads and performing segmentation extraction on the needle marks to compare the ratio of the needle marks on the test pads with a preset range includes: Respectively extracting the width, height and area data of the test pad and the needle marks; Respectively comparing the relationship between the ratio of the width of the needle mark to the width of the test pad and a first preset range, the ratio of the height of the needle mark to the height of the test pad and a second preset range, and the ratio of the area of the needle mark to the area of the test pad and a third preset range; When the ratios of the width, height and area of the needle mark on the test pad are respectively within the corresponding preset ranges, determining that the needle mark is a normal needle mark; When any one of the ratios of the width, height and area of the needle mark on the test pad is outside the corresponding preset range, determining that the needle mark is an abnormal needle mark.
4. The method for detecting abnormal needle marks of WAT according to claim 3, wherein, When a complete rectangle cannot be detected on the test pad or there are no needle marks on the test pad, determining that the type of the abnormal needle mark is needle mark offset; When any one of the ratios of the width, height and area of the needle mark on the test pad is greater than the maximum value of the corresponding preset range, determining that the type of the abnormal needle mark is needle mark too large; When any one of the ratios of the width, height and area of the needle mark on the test pad is less than the minimum value of the corresponding preset range, determining that the type of the abnormal needle mark is needle mark too shallow.
5. The method for detecting abnormal needle marks of WAT according to claim 3, characterized in that, The first preset range includes: 0.05 < W / W b <0.3; where W is the width of the needle mark, and W b is the width of the test pad; The second preset range includes: 0.05 < H / H b <0.8; where H is the height of the needle mark, and H b is the height of the test pad; The third preset range includes: 0.05 < S / S b < 0.3; where S is the height of the needle mark, and S b is the area of the test pad.
6. The method for detecting abnormal needle marks of WAT according to claim 1, wherein, The obtaining the closed contour data of the test pads on the acupuncture image includes: Successively performing edge detection, contour dilation and contour detection on the acupuncture image through digital image processing algorithms to obtain the closed contour data of the test pads on the acupuncture image.
7. The method for detecting abnormal needle marks of WAT according to claim 1, characterized in that The locating all the test pads on the acupuncture image based on the closed contour data includes: Cleaning the test pad coordinates from the closed contour data; Based on the test pad coordinates, obtaining the width, height and area of the test pad and forming an aspect ratio range and an area range; Based on the aspect ratio range and the area range, all the test pads are located on the needle insertion image.
8. The method for detecting abnormal stitch marks of WAT according to claim 7, wherein, The aspect ratio range includes: 4 / 3 ≤ H b / W b ≤ 3 / 2; where H b is the height of the test pad, and W b is the width of the test pad; The area range includes: 480 ≤ S b ≤ 520, where S b is the area of the test pad.
9. The method for detecting abnormal needle marks of WAT according to claim 3 or 7, characterized in that Extract the width and height of the needle mark or the test pad, including: Extract the maximum and minimum values in the horizontal direction and the maximum and minimum values in the vertical direction from the set of contour points respectively; Take the difference between the maximum and minimum values in the horizontal direction as the width; take the difference between the maximum and minimum values in the vertical direction as the height.
10. A system for detecting abnormal needle marks in WAT, characterized in that, Applied to the method for detecting abnormal needle marks of WAT as described in any one of claims 1 to 8, including: An acquisition unit for acquiring a needle insertion image; A preprocessing unit communicatively connected to the acquisition unit for obtaining the closed contour data of the test pads on the needle insertion image; A positioning unit communicatively connected to the preprocessing unit for positioning all the test pads on the needle insertion image based on the closed contour data; A processing unit communicatively connected to the positioning unit for performing rectangular detection on the test pads, extracting the size data of the test pads, and comparing the ratio of the needle mark on the test pad with a preset range to determine whether the needle mark is normal and classifying the abnormal needle marks.