Needle Mark Filter Detection Method, Electronic Device and Storage Medium

Abstract

The present application discloses a method for detecting and filtering needle marks, an electronic device, and a storage medium. The method for detecting and filtering needle marks includes: obtaining an image of a pad area of a chip to be detected; determining an initial binary image based on the image of the pad area; determining a target binary image based on the initial binary image and a noise contour threshold; determining a set of needle mark distribution images based on the target binary image; determining the cell position corresponding to each needle mark to be detected based on the set of needle mark distribution images; and determining whether there is a needle mark to be filtered in the cell position corresponding to each needle mark to be detected based on a needle mark template image and the cell position corresponding to each needle mark to be detected. By using the technical solution of the present application, the position of the needle mark can be efficiently and accurately located, the misjudgment rate of the needle mark can be reduced, the occurrence of over-inspection can be reduced, and the accuracy and robustness of chip detection can be improved.

Description

Technical Field

[0001] This application generally relates to the field of image processing technology. More specifically, this application relates to a method for filtering and detecting pin marks, an electronic device, and a storage medium. Background Art

[0002] During the production process of SIC (silicon carbide) semiconductor chips, there may be a large number of pin marks in the Pad (pin pad) area of the chips. The reasons may be that the probes are worn and dull due to long use time or excessive test times, resulting in uneven pressure distribution when the probes contact the Pad, forming irregular pin marks. It may also be due to poor probe tip process, with problems such as irregular shape or burrs, so that abnormal pin marks are left when the probes contact the Pad. It may also be that the perpendicularity is not enough when the probes contact the Pad, which will cause the contact points of the probes and the Pad to deviate from the center position, resulting in pin mark deviation or irregular shape. It may also be that the moving accuracy of the carrier stage of the pin tester is not high, and the wafer cannot be accurately moved to the relative position of the probe card, resulting in inaccurate positions when the probes contact the Pad and pin mark deviation. Excessive pin marks will lead to over-inspection during the AOI (Automated Optical Inspection) defect detection of the Pad area of the chips.

[0003] The existing methods for filtering pin marks are generally of two types. One type is the method of using manual shielding. For example, when building the AOI model, all pin mark areas are deducted on the model layout and no corresponding detection is performed, thus avoiding over-inspection and correspondingly reducing the subsequent re-judgment. However, although this method can reduce over-inspection, there is a risk of missed inspection. Once the pin marks deviate, the filtering will be ineffective. The other type is the method of using supervised deep learning, such as anomaly detection based on VAND to filter pin marks, but it takes a lot of time and labor costs to collect a large number of images and label the pin marks, and it also takes a lot of time to train the model so that the model can learn the ability to detect pin marks.

[0004] In view of this, there is an urgent need to provide a method for filtering and detecting pin marks, so as to be able to efficiently and accurately locate the positions of pin marks, reduce the misjudgment rate of pin marks, reduce the occurrence of over-inspection, and improve the accuracy and robustness of chip detection. Summary of the Invention

[0005] In order to solve at least one or more of the above-mentioned technical problems, this application proposes a method for filtering and detecting pin marks, an electronic device, and a storage medium in multiple aspects. This method for filtering and detecting pin marks can efficiently and accurately locate the positions of pin marks, reduce the misjudgment rate of pin marks, reduce the occurrence of over-inspection, and improve the accuracy and robustness of chip detection.

[0006] In a first aspect, the present application provides a method for detecting and filtering needle marks, including: obtaining an image of the pad area of a chip to be detected; determining an initial binary image based on the image of the pad area; determining a target binary image based on the initial binary image and a noise contour threshold; determining a set of needle mark distribution images based on the target binary image; determining the cell position corresponding to each needle mark to be detected based on the set of needle mark distribution images; and determining whether there are needle marks to be filtered in the cell position corresponding to each needle mark to be detected based on a needle mark template image and the cell position corresponding to each needle mark to be detected.

[0007] In some embodiments, determining the initial binary image based on the image of the pad area includes: converting the image of the pad area into a grayscale image; determining a segmentation threshold based on the grayscale image; and performing threshold segmentation on the grayscale image according to the segmentation threshold to obtain the initial binary image.

[0008] In some embodiments, determining the target binary image based on the initial binary image and the noise contour threshold includes: setting the noise contour threshold; comparing the contour size of each closed region contour in the initial binary image with the noise contour threshold, and determining the target binary image according to the comparison result.

[0009] In some embodiments, determining the set of needle mark distribution images based on the target binary image includes: performing dilation processing on the target binary image based on a first rectangular convolution kernel to obtain a dilated image; performing erosion processing on the dilated image based on a second rectangular convolution kernel to obtain a closing operation image; and determining the set of needle mark distribution images based on the closing operation image.

[0010] In some embodiments, determining the set of needle mark distribution images based on the closing operation image includes: respectively extracting the horizontal needle mark distribution images corresponding to each pixel column in the closing operation image, and respectively extracting the vertical needle mark distribution images corresponding to each pixel row in the closing operation image; and constituting the set of needle mark distribution images based on the horizontal needle mark distribution images corresponding to each pixel column and the vertical needle mark distribution images corresponding to each pixel row.

[0011] In some embodiments, determining the cell position corresponding to each needle mark to be detected based on the set of needle mark distribution images includes: determining the distribution points corresponding to each needle mark to be detected according to the horizontal needle mark distribution images corresponding to each pixel column and the vertical needle mark distribution images corresponding to each pixel row; and determining the cell position corresponding to each needle mark to be detected according to the distribution points corresponding to each needle mark to be detected and the preset distribution positions corresponding to each cell.

[0012] In some embodiments, determining whether there are needle marks to be filtered in the cell position corresponding to each needle mark to be inspected based on the needle mark template image and the cell position corresponding to each needle mark to be inspected includes: comparing the needle marks to be inspected in the cell position corresponding to each needle mark to be inspected with the template needle marks in each template cell in the needle mark template image for feature difference comparison; and determining whether there are needle marks to be filtered in the cell position corresponding to each needle mark to be inspected based on the feature difference comparison result.

[0013] In some embodiments, after determining whether there are needle marks to be filtered in the cell position corresponding to each needle mark to be inspected based on the feature difference comparison result, the method further includes: if it is determined that there are needle marks to be filtered in the cell position corresponding to the needle mark to be inspected, then subtracting the contour of the needle mark to be inspected in the cell position corresponding to the needle mark to be inspected from the contour of the template needle mark in the template cell corresponding to the current cell position to obtain a filtered needle mark contour; and respectively marking each filtered needle mark contour to filter each marked filtered needle mark contour when performing AOI defect detection on the pad area of the chip to be detected.

[0014] In a second aspect, the present application provides an electronic device, including: a processor; and a memory, on which program code for needle mark filtering detection is stored, and when the program code is executed by the processor, the electronic device implements the method as described above.

[0015] In a third aspect, the present application provides a non-transitory machine-readable storage medium, on which program code for needle mark filtering detection is stored, and when the program code is executed by a processor, the method as described above can be implemented.

[0016] The technical solution provided by the present application may include the following beneficial effects:

[0017] The needle mark filtering detection method, electronic device and storage medium provided by the present application, by acquiring the pad area image of the chip to be detected, and then determining the initial binary image based on the pad area image, and then determining the target binary image based on the initial binary image and the noise contour threshold, is beneficial to removing interference noise and reducing the false judgment rate of needle marks. Furthermore, determining the needle mark distribution image set based on the target binary image can clarify the distribution of each needle mark to be inspected on the pad area of the chip to be detected and accurately locate the position of the needle mark on the pad area of the chip to be detected.

[0018] Further, the present application can determine the cell position corresponding to each needle mark to be detected based on the set of needle mark distribution images, and then can determine whether there are needle marks to be filtered in the cell position corresponding to each needle mark to be detected based on the needle mark template image and the cell position corresponding to each needle mark to be detected. Thus, irregular needle marks and offset needle marks that need to be filtered during AOI defect detection of the pad area of the chip to be detected can be detected, reducing the occurrence of over-inspection in AOI defect detection and improving the accuracy and robustness of chip detection.

[0019] Generally speaking, the present application can efficiently and accurately locate the needle mark position, reduce the misjudgment rate of needle marks, reduce the occurrence of over-inspection, and improve the accuracy and robustness of chip detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present application will become readily understood. In the drawings, several embodiments of the present application are shown in an exemplary rather than restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0021] Figure 1 Shows an exemplary flowchart of a needle mark filtering detection method according to some embodiments of the present application;

[0022] Figure 2 Shows an exemplary flowchart of a needle mark filtering detection method according to other embodiments of the present application;

[0023] Figure 3 Shows an exemplary flowchart of a needle mark filtering detection method according to still other embodiments of the present application;

[0024] Figure 4 Shows an example diagram of an initial binary image in the needle mark filtering detection method according to an embodiment of the present application;

[0025] Figure 5 Shows an example diagram of a target binary image in the needle mark filtering detection method according to an embodiment of the present application;

[0026] Figure 6 Shows an example diagram of a grayscale image in the needle mark filtering detection method according to an embodiment of the present application;

[0027] Figure 7 Shows an example diagram of a dilated image in the needle mark filtering detection method according to an embodiment of the present application;

[0028] Figure 8 Shows an example diagram of a closing operation image in the needle mark filtering detection method according to an embodiment of the present application;

[0029] Figure 9Shows an example diagram of the horizontal needle mark distribution image in the needle mark filtering detection method according to an embodiment of the present application;

[0030] Figure 10 Shows an example diagram of the vertical needle mark distribution image in the needle mark filtering detection method according to an embodiment of the present application;

[0031] Figure 11 Shows an example diagram formed after determining the cell positions corresponding to each needle mark to be detected in the needle mark filtering detection method according to an embodiment of the present application;

[0032] Figure 12 Shows a schematic structural diagram of an electronic device according to an embodiment of the present application. Detailed implementation manners

[0033] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. For the sake of simplicity and clarity of description, where appropriate, the same reference numerals may be repeated in the drawings to indicate corresponding or similar elements. In addition, the present application elaborates on many specific details to provide a thorough understanding of the embodiments described herein. However, those of ordinary skill in the art will understand that the embodiments described herein can be practiced without these specific details. In other cases, well-known methods, processes, and components are not described in detail so as not to obscure the embodiments described herein. Moreover, this description should not be regarded as limiting the scope of the embodiments described herein. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts fall within the scope of protection of the present application.

[0034] It should be understood that the possible terms "first" or "second" etc. in the claims, the description, and the drawings disclosed in the present application are used to distinguish different objects, rather than to describe a specific order. The terms "including" and "comprising" used in the description and claims of the present application indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0035] It should also be understood that the terms used in the description of the present application herein are only for the purpose of describing specific embodiments, and are not intended to limit the present application. As used in the description and claims of the present application, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms. It should be further understood that the term " / and" used in the description and claims of the present application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0036] As used in this specification and the claims, the term "if" can be interpreted as "when", "once", "in response to determining", or "in response to detecting" depending on the context. Similarly, the phrases "if determined" or "if [the described condition or event] is detected" can be interpreted as meaning "once determined", "in response to determining", "once [the described condition or event] is detected", or "in response to detecting [the described condition or event]" depending on the context.

[0037] During the production process of SIC (silicon carbide) semiconductor chips, there may be a large number of pin marks in the Pad area of the chips. The reasons may be that the probes are worn and dull due to long usage time or excessive testing times, resulting in uneven pressure distribution when the probes contact the Pad, forming irregular pin marks. It may also be due to poor probe tip process, with problems such as irregular shape or burrs, so that abnormal pin marks are left when the probes contact the Pad. It may also be that the perpendicularity is insufficient when the probes contact the Pad, which will cause the contact points of the probes and the Pad to deviate from the center position, resulting in pin mark offset or irregular shape. It may also be that the moving accuracy of the carrier stage of the pin tester is not high, and the wafer cannot be accurately moved to the relative position of the probe card, resulting in inaccurate positions when the probes contact the Pad and pin mark offset. Excessive pin marks will cause over-inspection during AOI (Automated Optical Inspection) defect detection of the Pad area of the chips. The existing methods for filtering pin marks are generally of two types. One type is the method of manual shielding. Although it can reduce over-inspection, there is a risk of missed inspection. Once the pin marks deviate, the filtering is ineffective. The other type is the method of supervised deep learning, but it takes a lot of time and labor costs to collect a large number of images and label the pin marks, and it also takes a lot of time to train the model so that the model learns the ability to detect pin marks.

[0038] In view of this, there is an urgent need to provide a method for filtering and detecting pin marks, so as to be able to efficiently and accurately locate the positions of pin marks, reduce the misjudgment rate of pin marks, reduce the occurrence of over-inspection, and improve the accuracy and robustness of chip detection.

[0039] The following will describe in detail the specific embodiments of the present application with reference to the accompanying drawings.

[0040] Figure 1 Exemplary flowchart 100 of the pin mark filtering and detecting method according to some embodiments of the present application is shown. Figure 4 An example diagram of the initial binary image in the pin mark filtering and detecting method according to an embodiment of the present application is shown. Figure 5 An example diagram of the target binary image in the pin mark filtering and detecting method according to an embodiment of the present application is shown. Figure 11The figure shows an example diagram formed after determining the cell positions corresponding to each needle mark to be detected in the needle mark filtering detection method according to an embodiment of the present application. Please refer to Figure 1 , Figure 4 , Figure 5 and Figure 11 . The needle mark filtering detection method shown in the embodiments of the present application may include:

[0041] In step S101, an image of the pad area of the chip to be detected is acquired. In the embodiments of the present application, the chip to be detected may be a SIC (silicon carbide) semiconductor chip. The pad area (Pad area) of the chip to be detected refers to the area of the layout connection points or pins on the chip to be detected, which is used to connect the internal circuit of the chip to be detected with the external circuit.

[0042] In addition, in the embodiments of the present application, a high-resolution industrial camera may be used to collect the image of the pad area. It can be understood that the method of acquiring the image of the pad area is diverse. In practical applications, a suitable method needs to be selected according to the actual application situation to acquire the image of the pad area, and the present application does not impose any restrictions in this regard.

[0043] In step S102, an initial binary image is determined based on the image of the pad area. As Figure 4 shown, in the embodiments of the present application, the image of the pad area may be subjected to image binarization processing. Image binarization refers to the process of converting a color or grayscale image into a binary image (also called a black-and-white image). A binary image has only two pixel values, generally 0 and 255, representing black and white respectively, so as to obtain an initial binary image.

[0044] In step S103, a target binary image is determined based on the initial binary image and the noise contour threshold. As Figure 4 shown, there are still quite a few interference noises in the initial binary image obtained after image binarization, which are likely to interfere with the determination of the needle mark position. Therefore, in the embodiments of the present application, it is necessary to segment and remove the interference noises through a preset noise contour threshold to obtain the target binary image as Figure 5 shown.

[0045] In step S104, a needle mark distribution image set is determined based on the target binary image. In the embodiments of the present application, the target binary image may be sequentially segmented into a plurality of column pixel columns and a plurality of row pixel rows. It can be understood that each pixel column obtained by sequentially segmenting the target binary image can form a pixel column image. Similarly, each pixel row obtained by sequentially segmenting the target binary image can form a pixel row image. The aforementioned obtained plurality of images can constitute a needle mark distribution image set.

[0046] In step S105, based on the set of pin mark distribution images, determine the cell position corresponding to each pin mark to be inspected. In the embodiments of the present application, the distribution of pin mark highlights on each pixel column and each pixel row can be determined by using several pixel column images and several pixel row images in the above set of pin mark distribution images. Thus, the distribution of each pin mark highlight on the target binary image can be determined by using the distribution of pin mark highlights on each pixel column and each pixel row. The information of the distribution can include but is not limited to the highlight area, highlight coordinates, etc. of each pin mark highlight. This is equivalent to clarifying the distribution of each pin mark to be inspected on the pad area of the chip to be detected.

[0047] It can be understood that in the embodiments of the present application, several cells can be pre - laid out in the grayscale image. Furthermore, the cell where each pin mark to be inspected is located can be determined according to the distribution of each pin mark to be inspected. For example, the coordinate information of the current pin mark to be inspected (such as the highlight center coordinate of the pin mark highlight) can be used to determine whether the current pin mark to be inspected is within the coordinate range of the current cell. If it is, it is determined that the current pin mark to be inspected is within the current cell, so that the cell position corresponding to each pin mark to be inspected can be determined.

[0048] In step S106, based on the pin mark template image and the cell position corresponding to each pin mark to be inspected, determine whether there are pin marks to be filtered in the cell position corresponding to each pin mark to be inspected. In the embodiments of the present application, the pin mark template image can also be laid out with template cells, and a template pin mark contour can also be set in each template cell. It can be understood that the positions of the several cells pre - laid out in the grayscale image should correspond one - to - one with the positions of the template cells in the pin mark template image. In this way, the pin marks to be inspected in the cell corresponding to each pin mark to be inspected can be compared one by one with the template pin marks in each template cell in the pin mark template image according to the cell position corresponding to each pin mark to be inspected. Thus, it can be determined whether there are pin marks to be filtered in the cell position corresponding to each pin mark to be inspected according to the comparison result. Among them, the pin marks to be filtered refer to the pin marks that need to be filtered when performing AOI defect detection on the pad area of the chip to be detected, and can include but are not limited to pin marks with irregular shapes and pin marks with position offsets, etc.

[0049] In the embodiments of the present application, by obtaining the image of the pad area of the chip to be detected, and then determining the initial binary image based on the pad area image, and further determining the target binary image based on the initial binary image and the noise contour threshold, it is beneficial to remove interference noise and reduce the misjudgment rate of pin marks. Then, based on the target binary image, a set of pin mark distribution images is determined, so as to clarify the distribution of each pin mark to be detected on the pad area of the chip to be detected and accurately locate the position of the pin mark on the pad area of the chip to be detected. Further, the present application can determine the cell position corresponding to each pin mark to be detected based on the set of pin mark distribution images, and then can determine whether there are pin marks to be filtered in the cell position corresponding to each pin mark to be detected based on the pin mark template image and the cell position corresponding to each pin mark to be detected. Thus, irregular pin marks and offset pin marks that need to be filtered during AOI defect detection of the pad area of the chip to be detected can be detected, reducing the occurrence of over-inspection in AOI defect detection and improving the accuracy and robustness of chip detection. Generally speaking, the present application can efficiently and accurately locate the position of pin marks, reduce the misjudgment rate of pin marks, reduce the occurrence of over-inspection, and improve the accuracy and robustness of chip detection.

[0050] In some embodiments, the determination process of the set of pin mark distribution images can be further designed. Figure 2 FIG. shows an exemplary flowchart of the pin mark filtering and detection method according to other embodiments of the present application. Figure 6 FIG. shows an example diagram of the grayscale image in the pin mark filtering and detection method according to the embodiments of the present application. Figure 7 FIG. shows an example diagram of the dilated image in the pin mark filtering and detection method according to the embodiments of the present application. Figure 8 FIG. shows an example diagram of the closing operation image in the pin mark filtering and detection method according to the embodiments of the present application. Figure 9 FIG. shows an example diagram of the horizontal pin mark distribution image in the pin mark filtering and detection method according to the embodiments of the present application. Figure 10 FIG. shows an example diagram of the vertical pin mark distribution image in the pin mark filtering and detection method according to the embodiments of the present application. Please refer to Figure 2 and Figures 4 to 10 , the pin mark filtering and detection method shown in the embodiments of the present application may include:

[0051] In step S201, an initial binary image is determined based on the pad area image. In the embodiments of the present application, the obtained pad area image can be first converted into a grayscale image as shown in Figure 6 . A color image usually consists of three color channels: red (R), green (G), and blue (B). The pixel value range of each channel is also 0 to 255. The purpose of grayscale conversion is to combine the values of these three channels into a single grayscale value, which is beneficial to reducing the data volume, simplifying the image processing process, and at the same time retaining the main structural information of the image.

[0052] Then, the segmentation threshold can be determined based on the grayscale image. In the embodiments of the present application, the segmentation threshold can be determined by the following formula (1):

[0053]

[0054] Wherein, is the lower limit value of the optimal threshold and can be used as the above-mentioned segmentation threshold, is the pixel coordinate of the image of the pad area. Furthermore, the grayscale image can be threshold-segmented according to the segmentation threshold to obtain an initial binary image as shown in Figure 4 .

[0055] In step S202, a target binary image is determined based on the initial binary image and the noise contour threshold. In the embodiments of the present application, the noise contour threshold can be set first. Exemplarily, the noise contour threshold can be obtained by statistically analyzing the size data of historical noise contours. For example, the average value or standard deviation obtained by statistics can be used as the noise contour threshold. It can be understood that the setting method of the noise contour threshold is diverse. In practical applications, the noise contour threshold needs to be reasonably set according to the actual application situation, and the present application does not impose any restrictions in this regard.

[0056] Then, the contour size of each closed region contour in the initial binary image can be compared with the noise contour threshold, and the target binary image can be determined according to the comparison result. Specifically, the pixel values corresponding to the closed region contours with contour sizes smaller than the noise contour threshold can be updated to 0, so as to obtain a target binary image as shown in Figure 5 . Exemplarily, the comparison can be performed by the following formula (2):

[0057]

[0058] Wherein, is the target binary image, is the number of the closed region contour, is the noise contour threshold.

[0059] In step S203, the target binary image is dilated based on the first rectangular convolution kernel to obtain a dilated image. The aforementioned first rectangular convolution kernel can be a 3×3 rectangular convolution kernel, and all the element values in the convolution kernel are 1. In the embodiments of the present application, the dilation processing can be exemplarily performed by the following formula (3) to obtain a dilated image as shown in Figure 7 :

[0060]

[0061] Wherein, is the dilated image, is the target binary image, is the first rectangular convolution kernel. is the row number in the target binary image, is the column number in the target binary image.

[0062] In step S204, the dilated image is eroded based on the second rectangular convolution kernel to obtain a closing operation image. The aforementioned second rectangular convolution kernel can be a 3x3 rectangular convolution kernel, and all element values in the convolution kernel are 1. In the embodiments of the present application, the erosion process can be exemplarily performed through the following formula four to obtain the closing operation image as Figure 8 shown:

[0063]

[0064] where, is the closing operation image obtained after eroding the dilated image, is the second rectangular convolution kernel.

[0065] In step S205, the stitch mark distribution image set is determined based on the closing operation image. In the embodiments of the application, as Figure 9 and Figure 10 shown, the closing operation image can be sequentially segmented, and the horizontal stitch mark distribution image corresponding to each pixel column in the closing operation image can be extracted respectively. It can be exemplarily extracted through the following formula five:

[0066]

[0067] where, is the horizontal stitch mark distribution image corresponding to the nth pixel column, is the nth column in the closing operation image.

[0068] And, the vertical stitch mark distribution image corresponding to each pixel row in the closing operation image is extracted respectively. It can be exemplarily extracted through the following formula six:

[0069]

[0070] where, is the vertical stitch mark distribution image corresponding to the mth pixel row, is the mth row in the closing operation image. Thus, the stitch mark distribution image set is constituted based on the horizontal stitch mark distribution image corresponding to each pixel column and the vertical stitch mark distribution image corresponding to each pixel row.

[0071] In some embodiments, the outline of the filtered stitch marks can be marked after determining whether there are stitch marks to be filtered in the cell position corresponding to each stitch mark to be detected. Figure 3 shows an exemplary flowchart of the stitch mark filtering and detection method according to still some embodiments of the present application. Please refer to Figure 3And Figures 9 to 11 The pin mark filtering and detection method shown in the embodiments of this application may include:

[0072] In step S301, based on the set of pin mark distribution images, determine the cell position corresponding to each pin mark to be detected. In the embodiments of this application, the distribution points corresponding to each pin mark to be detected can be determined according to the horizontal pin mark distribution image corresponding to each pixel column and the vertical pin mark distribution image corresponding to each pixel row ( Figure 9 And Figure 10 The white line segments in are the distribution points of the pin marks to be detected). Further, the cell position corresponding to each pin mark to be detected can be determined according to the distribution points corresponding to each pin mark to be detected and the preset distribution position corresponding to each cell. Specifically, a number of cells can be pre-arranged in the grayscale image, and then the cell where each pin mark to be detected is located can be determined according to the distribution of each pin mark to be detected. For example, the coordinate information of the current pin mark to be detected (such as the center coordinate of the bright point of the pin mark) can be used to determine whether the current pin mark to be detected is within the coordinate range of the current cell. If it is, it is determined that the current pin mark to be detected is within the current cell, so that the cell position corresponding to each pin mark to be detected can be determined.

[0073] In step S302, based on the pin mark template image and the cell position corresponding to each pin mark to be detected, determine whether there are pin marks to be filtered in the cell position corresponding to each pin mark to be detected. In the embodiments of this application, as Figure 11 shown, the pin marks to be detected in the cell position corresponding to each pin mark to be detected can be compared with the template pin marks in each template cell in the pin mark template image for feature difference comparison. The feature difference comparison may include but is not limited to the comparison of the number of pin marks and the comparison of the pin mark shapes, etc. Among them, the comparison of the pin mark shapes can be carried out by extracting the contour of the pin mark to be detected in the current cell position and extracting the contour of the template pin mark in the template cell corresponding to the current cell position. The extraction method can use the cv2.findContours function in OpenCV, and the comparison method can use the matchShape function. The matchShape function can calculate the distance between two contours and return a value. The smaller the value, the more similar the two contours are.

[0074] Based on the feature difference comparison result, it is determined whether there are needle marks to be filtered in the cell position corresponding to each detected needle mark. It can be understood that if the number of detected needle marks in the current cell position is more than the number of template needle marks in the template cell corresponding to the current cell position, it indicates that there are needle marks to be filtered in the current cell position. If the similarity between the contour of the detected needle mark in the current cell position and the contour of the template needle mark in the template cell corresponding to the current cell position is less than the preset similarity (exemplarily, the preset similarity can be set to 70%, and the present application makes no restrictions), it indicates that there are needle marks to be filtered in the current cell position.

[0075] In step S303, if it is determined that there are needle marks to be filtered in the cell position corresponding to the detected needle mark, the filtered needle mark contour is determined. In the embodiment of the present application, if it is determined that there are needle marks to be filtered in the cell position corresponding to the detected needle mark, the contour of the detected needle mark in the cell position corresponding to the detected needle mark is subtracted from the contour of the template needle mark in the template cell corresponding to the current cell position to obtain the filtered needle mark contour.

[0076] In step S304, each filtered needle mark contour is marked respectively. The marking method can be to frame the contour edge of the filtered needle mark contour and set a marking label, so as to filter each marked filtered needle mark contour when performing AOI defect detection on the pad area of the chip to be detected.

[0077] Corresponding to the foregoing application function implementation method embodiment, the present application also provides an electronic device for executing the needle mark filtering detection method and a corresponding embodiment.

[0078] Figure 12 The block diagram showing the hardware configuration of the electronic device 1200 that can implement the needle mark filtering detection method of the embodiment of the present application. As Figure 12 shown, the electronic device 1200 may include a processor 1210 and a memory 1220. In Figure 12 the electronic device 1200, only the constituent elements related to this embodiment are shown. Therefore, it is obvious to those of ordinary skill in the art that: the electronic device 1200 may also include common constituent elements different from Figure 12 the constituent elements shown in. For example: a fixed-point arithmetic unit.

[0079] The electronic device 1200 may correspond to a computing device with various processing functions. For example, functions for generating a neural network, training or learning a neural network, quantifying a floating-point neural network into a fixed-point neural network, or retraining a neural network. For example, the electronic device 1200 may be implemented as various types of devices, such as a personal computer (PC), a server device, a mobile device, etc.

[0080] The processor 1210 controls all functions of the electronic device 1200. For example, the processor 1210 controls all functions of the electronic device 1200 by executing programs stored in the memory 1220 of the electronic device 1200. The processor 1210 may be implemented by a central processing unit (CPU), a graphics processing unit (GPU), an application processor (AP), an artificial intelligence processor chip (IPU), etc. provided in the electronic device 1200. However, the present application is not limited thereto.

[0081] In some embodiments, the processor 1210 may include an input / output (I / O) unit 1211 and a computing unit 1212. The I / O unit 1211 may be used to receive various data, such as an image of the pad area of the chip to be detected. Exemplarily, the computing unit 1212 may be used to determine an initial binary image based on the pad area image received via the I / O unit 1211; determine a target binary image based on the initial binary image and a noise contour threshold; determine a set of stitch mark distribution images based on the target binary image; determine the cell position corresponding to each stitch mark to be detected based on the set of stitch mark distribution images; and determine whether there is a stitch mark to be filtered in the cell position corresponding to each stitch mark to be detected based on the stitch mark template image and the cell position corresponding to each stitch mark to be detected. The result of whether there is a stitch mark to be filtered may be output by the I / O unit 1211, for example. The output data may be provided to the memory 1220 for other devices (not shown) to read and use, or may be directly provided to other devices for use.

[0082] The memory 1220 is hardware for storing various data processed in the electronic device 1200. For example, the memory 1220 can store the processed data and the data to be processed in the electronic device 1200. The memory 1220 can store the data sets involved in the process of the pin mark filtering detection method that the processor 1210 has processed or is to process, such as the pad area image, etc. In addition, the memory 1220 can store the applications, drivers, etc. to be driven by the electronic device 1200. For example, the memory 1220 can store various programs related to the pin mark filtering detection method to be executed by the processor 1210. The memory 1220 can be a DRAM, but the present application is not limited thereto. The memory 1220 can include at least one of a volatile memory or a non-volatile memory. The non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, phase change RAM (PRAM), magnetic RAM (MRAM), resistive RAM (RRAM), ferroelectric RAM (FRAM), etc. The volatile memory can include dynamic RAM (DRAM), static RAM (SRAM), synchronous DRAM (SDRAM), PRAM, MRAM, RRAM, ferroelectric RAM (FeRAM), etc. In an embodiment, the memory 1220 can include at least one of a hard disk drive (HDD), a solid state drive (SSD), a high density flash (CF), a secure digital (SD) card, a micro secure digital (Micro-SD) card, a mini secure digital (Mini-SD) card, an extreme digital (xD) card, caches, or a memory stick.

[0083] In summary, the specific functions implemented by the memory 1220 and the processor 1210 of the electronic device 1200 provided in the embodiments of this specification can be explained in contrast to the foregoing embodiments in this specification, and can achieve the technical effects of the foregoing embodiments, which will not be elaborated here.

[0084] In this embodiment, the processor 1210 can be implemented in any suitable manner. For example, the processor 1210 can take the form of, for example, a microprocessor or a processor and a computer-readable medium storing computer-readable program code (such as software or firmware) executable by the (micro)processor, logic gates, switches, an application specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller, and so on.

[0085] It should also be understood that any module, unit, component, server, computer, terminal, or device that executes instructions as exemplified herein may include or otherwise access a computer-readable medium, such as a storage medium, a computer storage medium, or a data storage device (removable) and / or non-removable), such as a magnetic disk, an optical disk, or a magnetic tape. A computer storage medium may include volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable instructions, data structures, program modules, or other data.

[0086] Although several embodiments of the present application have been shown and described herein, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many changes, alterations, and alternative ways will occur to those skilled in the art without departing from the spirit and scope of the present application. It should be understood that various alternatives to the embodiments of the present application described herein may be employed in practicing the present application. The appended claims are intended to define the scope of the present application and thus cover equivalents or alternatives within the scope of these claims.

Claims

1. A needle mark filtration detection method, characterized in that: include: Acquire a pad area image of the chip to be inspected; Determine an initial binary image based on the pad area image; Determine a target binary image based on the initial binary image and a noise contour threshold; Determine a needle mark distribution image set based on the target binary image; Determine the cell position corresponding to each needle mark to be detected based on the needle mark distribution image set; Determine whether there is a needle mark to be filtered in the cell position corresponding to each needle mark to be detected based on the needle mark template image and the cell position corresponding to each needle mark to be detected; The step of determining a needle mark distribution image set based on the target binary image comprises: Performing dilation processing on the target binary image based on a first rectangular convolution kernel to obtain a dilated image; Performing corrosion processing on the dilated image based on a second rectangular convolution kernel to obtain a closed operation image; The needle mark distribution image set is determined based on the closed operation image.

2. The needle mark filtration detection method according to claim 1, characterized in that: Determining the initial binary image based on the pad area image comprises: Converting the pad area image into a grayscale image; determining a segmentation threshold based on the grayscale image; Threshold segmentation is performed on the grayscale image according to the segmentation threshold to obtain the initial binary image.

3. The needle mark filtration detection method according to claim 1, characterized in that: Determining the target binary image based on the initial binary image and the noise contour threshold comprises: Setting the noise profile threshold; The contour size of each closed area contour in the initial binary image is compared with the noise contour threshold, and the target binary image is determined according to the comparison result.

4. The needle mark filtration detection method according to claim 1, characterized in that: The determining the needle mark distribution image set based on the closed operation image comprises: Extracting a horizontal needle mark distribution image corresponding to each pixel column in the closed operation image, and extracting a vertical needle mark distribution image corresponding to each pixel row in the closed operation image; The needle mark distribution image set is formed based on the horizontal needle mark distribution image corresponding to each pixel column and the vertical needle mark distribution image corresponding to each pixel row.

5. The needle mark filtration detection method according to claim 4, characterized in that: Determining the cell position corresponding to each needle mark to be detected based on the needle mark distribution image set includes: Determine the distribution point corresponding to each needle mark to be detected according to the horizontal needle mark distribution image corresponding to each pixel column and the vertical needle mark distribution image corresponding to each pixel row; The cell position corresponding to each needle mark to be detected is determined according to the distribution point position corresponding to each needle mark to be detected and the preset distribution position corresponding to each cell.

6. The needle mark filtration detection method according to claim 1, characterized in that: The method of determining whether there is a needle mark to be filtered in the cell position corresponding to each needle mark to be detected based on the needle mark template image and the cell position corresponding to each needle mark to be detected comprises: Compare the feature differences between the needle mark to be detected in the cell position corresponding to each needle mark to be detected and the template needle mark in each template cell in the needle mark template image; Based on the feature difference comparison results, determine whether there is a needle mark to be filtered in the cell position corresponding to each needle mark to be detected.

7. The needle mark filtration detection method according to claim 6, characterized in that: After determining whether there is a needle mark to be filtered in the cell position corresponding to each needle mark to be detected based on the feature difference comparison result, the method further includes: If it is determined that there is a needle mark to be filtered in the cell position corresponding to the needle mark to be detected, subtract the needle mark contour to be detected in the cell position corresponding to the needle mark to be detected from the template needle mark contour in the template cell corresponding to the current cell position to obtain a filtered needle mark contour; Each filter needle mark contour is marked respectively, so that each marked filter needle mark contour is filtered when performing AOI defect inspection on the pad area of ​​the chip to be inspected.

8. An electronic device, characterized in that: include: processor; as well as A memory having program codes for needle mark filtering detection stored therein, wherein when the program codes are executed by the processor, the electronic device implements the method as claimed in any one of claims 1 to 7.

9. A non-transitory machine-readable storage medium having stored thereon a program code for needle mark filtration detection, wherein when the program code is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images