Surface defect detection method and system for integrated circuit packaging process

Abstract

The invention discloses a surface defect detection method and system for an integrated circuit packaging process. The method comprises the following steps: acquiring a surface digital image of the integrated circuit packaging process; carrying out image preprocessing operation on the acquired surface digital image; performing feature extraction on the preprocessed surface image in the integrated circuit packaging process; performing defect type classification operation on the image after feature extraction according to defect features; a corresponding defect detection method is adopted for each type of defects, and effective separation of a normal background and a defect area is completed; determining a defect region of each defect type; and judging whether the defects meet industrial production requirements or not, and outputting a surface defect detection result. According to the method, the surface defect image in the packaging process of the integrated circuit is preprocessed and then subjected to defect detection, interference of factors such as noise and low contrast on subsequent defect detection is avoided, and the defect detection precision is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of circuit defect detection, and in particular to a surface defect detection method and system for an integrated circuit packaging process. Background Art

[0002] The packaging process is an important part of the integrated circuit industry. In order to ensure the yield rate of the product, the integrated circuit packaging process needs to be fully inspected. For the manufacturing process of the integrated circuit packaging process where the circuit has reached the nanometer level, in order to clearly present the surface defects in the integrated circuit packaging process and avoid image distortion after magnification, a high-power microscope is needed during the acquisition process. Usually, a complete integrated circuit needs to be divided into hundreds of areas for image acquisition. There are always small overlapping parts between the images of each area, and it is difficult to obtain a standard template. Secondly, there are many types of defects generated in the integrated circuit packaging process, such as short circuits, open circuits, blocked holes, broken holes, scratches, creases, indentations, spots, oxidation, ink, foreign matter and other defects, which greatly increase the difficulty of defect detection. With the continuous improvement of integration and the increase in the amount of detection tasks, the detection difficulty is greatly increased. Therefore, it is urgent to develop machine vision-related detection technologies to achieve automated detection. Summary of the invention

[0003] In order to overcome the defects and shortcomings of the prior art, the present invention provides a surface defect detection method and system for an integrated circuit packaging process. The present invention first performs a preprocessing operation on the surface defect image of the integrated circuit packaging process and then performs defect detection, thereby avoiding the interference of noise, low contrast and other factors on subsequent defect detection and improving the accuracy of defect detection.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] The present invention provides a surface defect detection method for an integrated circuit packaging process, comprising the following steps:

[0006] Collect surface digital images of integrated circuit packaging process;

[0007] Performing image preprocessing operations on the collected surface digital images;

[0008] Extract features from the pre-processed surface image of the integrated circuit packaging process;

[0009] According to the defect features, the defect type classification operation is performed on the image after feature extraction;

[0010] Use corresponding defect detection methods for each type of defect to effectively separate the normal background and defect area;

[0011] Determine the defect area for each defect type;

[0012] Determine whether the defects meet industrial production requirements and output surface defect detection results.

[0013] As a preferred technical solution, the feature extraction is performed on the surface image of the preprocessed integrated circuit packaging process, and specifically a regularization method is used to optimize and improve the low-rank embedding algorithm to extract the feature of the surface image of the preprocessed integrated circuit packaging process.

[0014] As a preferred technical solution, the defect type classification operation is performed on the image after feature extraction according to the defect features, specifically, the defect type classification operation is performed on the image after feature extraction based on a symplectic group classifier.

[0015] As a preferred technical solution, the defect type classification operation is performed on the image after feature extraction based on the symplectic group classifier, specifically including:

[0016] Map the image dataset to Sp(2n) and take the n-dimensional row vectors on Sp(2n) to form a set Specifically expressed as:

[0017]

[0018] Among them, the vector (I 1 ,I 2 ,…,I n ,) represents any image sample;

[0019] Select Q i ∈Sp(2n), convert Q into the corresponding symplectic matrix and find its singular values;

[0020] Q i Acting on the image dataset, it is expressed as:

[0021]

[0022] Perform singular value decomposition on the training image samples, take the first k largest singular values, and construct the discriminant function;

[0023] Solve the discriminant function and output the category of the image dataset.

[0024] As a preferred technical solution, the defect types include: manufacturing-related line width or line spacing related scale parameter defects, manufacturing-related hole related parameter defects, manufacturing-related irregularity defects, manufacturing-related unevenness defects, pollution-related color defects and trauma-related scratch defects.

[0025] As a preferred technical solution, the corresponding defect detection method is used for each type of defect to complete the effective separation of the normal background and the defect area, specifically including:

[0026] Analyze the characteristics of various defects and the connections and characteristics between defects;

[0027] Detect manufacturing-related line width or line spacing scale parameter defects, manufacturing-related hole parameter defects;

[0028] Based on the Jardon curve theorem, the region segmentation of traumatic scratches and manufacturing-related non-uniform defects is performed;

[0029] Regional segmentation of contamination-type color defects based on the Riemann manifold method;

[0030] The line edges of the preprocessed surface image are optimized, the constant cross-sectional curvature is calculated, and whether there are manufacturing irregular defects is determined based on the constant cross-sectional curvature.

[0031] As a preferred technical solution, the region segmentation of traumatic scratches and manufacturing-related uneven defects is performed based on the Jardon curve theorem, specifically including:

[0032] The preprocessed surface image of the integrated circuit packaging process is converted into an HSV color space image, and a lightness V component of the color space image is obtained;

[0033] Construct the segmentation curve of surface image based on Jordan curve theorem;

[0034] Regional segmentation of trauma-related scratches and manufacturing-related uneven defects based on segmentation curves;

[0035] Based on the segmented area, determine whether there are trauma-related scratches and manufacturing-related uneven defects.

[0036] As a preferred technical solution, the regional segmentation of pollution-type color defects is performed based on the Riemann manifold method, specifically including:

[0037] Obtain image background samples, and map the image background samples to the Riemann manifold space;

[0038] Constructing a similarity model between the image background sample and the image to be tested, and calculating the similarity between the image background sample and the image to be tested;

[0039] Use the principle of low-probability events to obtain the optimal threshold;

[0040] The area larger than the optimal threshold is judged as the background, and the area smaller than the optimal threshold is judged as a defect, thus completing the regional segmentation of contamination-type color defects.

[0041] As a preferred technical solution, optimizing the line edge of the preprocessed surface image specifically includes:

[0042] The preprocessed surface image is converted from RGB color space to Lab color space, and image processing is performed based on morphological methods to extract line edges in the L component of the Lab color space;

[0043] A fitting operation is performed based on the principle of numerical approximation to optimize the line edges extracted from the L component.

[0044] The present invention also provides a surface defect detection system for an integrated circuit packaging process, which is used to implement the surface defect detection method for the integrated circuit packaging process, and the system comprises: a microscopic imaging acquisition device, an image preprocessing module, a defect detection module, and a surface defect detection result output module;

[0045] The microscopic imaging acquisition device is used to acquire surface digital images of the integrated circuit packaging process;

[0046] The image preprocessing module is used to perform image preprocessing operations on the collected surface digital images;

[0047] The defect detection module is used to detect different types of defect images after preprocessing and determine the defect area, specifically including:

[0048] Extract features from the pre-processed surface image of the integrated circuit packaging process;

[0049] According to the defect features, the defect type classification operation is performed on the image after feature extraction;

[0050] Use corresponding defect detection methods for each type of defect to effectively separate the normal background and defect area;

[0051] Determine the defect area for each defect type;

[0052] The surface defect detection result output module is used to determine whether the defects meet the industrial production requirements and output the surface defect detection results.

[0053] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0054] (1) The present invention performs preprocessing operations on the surface defect images of the integrated circuit packaging process before performing defect detection, thereby avoiding interference of factors such as noise and low contrast on subsequent defect detection and improving the accuracy of defect detection.

[0055] (2) The present invention adopts the Jardon curve correlation theory and designs a segmentation curve suitable for integrated circuit images, which effectively solves the detection of two types of defects: trauma-related scratches and manufacturing-related unevenness.

[0056] (3) The present invention adopts the Riemannian manifold correlation theory and uses the Riemannian metric to effectively construct a similarity model between the background sample and the image to be tested, thereby achieving effective segmentation of contamination-type color defects.

[0057] (4) The present invention combines the cross-sectional curvature related theorem to design an algorithm suitable for manufacturing type irregular defect detection, which not only maintains the edge characteristics of the line to the greatest extent, but also improves the processing speed of the algorithm.

[0058] (5) The present invention improves the quality control of the diverse defects existing in the integrated circuit packaging process and improves the reliability of the integrated circuit packaging process. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a schematic diagram of the process of the surface defect detection system used in the integrated circuit packaging process of the present invention;

[0060] Figure 2 A schematic diagram of the implementation process of determining trauma-related scratches and manufacturing-related non-uniform defect areas according to the present invention;

[0061] Figure 3 A schematic diagram of the implementation process of determining the pollution type color defect area of ​​the present invention;

[0062] Figure 4 The present invention is a schematic diagram of the implementation process of determining the manufacturing type irregular defect area. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention 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 invention and are not intended to limit the present invention.

[0064] Example 1

[0065] like Figure 1 As shown, this embodiment provides a surface defect detection method for an integrated circuit packaging process, comprising the following steps:

[0066] S1: Collect digital images of the surface of the integrated circuit packaging process through high-magnification microscopy, including:

[0067] S11: Set the current position and related parameters of the microscope;

[0068] S12: planning the acquisition path of the loading platform to ensure the integrity of image acquisition;

[0069] In step S12, a complete integrated circuit needs to be divided into hundreds of areas for image acquisition. There is always a small overlap between the images of each area. Therefore, during the acquisition process, while ensuring the integrity of the image acquisition, it is also necessary to control the overlap of each area, thereby avoiding repeated operations in subsequent defect detection.

[0070] S13: The microscope collects the surface image of the integrated circuit packaging process. The microscope collects the RGB color space image;

[0071] S14: the stage moves to the next position, and collects the surface image of the current position after the stage is stable;

[0072] S15: the stage continues to move and determines whether the stage has moved to the set destination position;

[0073] S16: If the image acquisition is achieved, the image acquisition is completed; otherwise, the process returns to step S14 to continue the image acquisition operation.

[0074] S2: Perform image preprocessing on the acquired images;

[0075] In step S2, since high-magnification microscopic imaging is used to collect surface images of the integrated circuit packaging process, the complexity of the environment can easily lead to poor image quality. For example, a variety of unknown noises are introduced in the industrial environment, the image is dark overall, the contrast is low, etc. No matter what factors affect the image and cause the quality to deteriorate, the effect of subsequent defect detection will be reduced. Only by preprocessing the collected image can it be more helpful to improve the accuracy of defect detection.

[0076] In this embodiment, the image preprocessing operation includes but is not limited to image preprocessing such as noise reduction, improving the overall brightness of the image, and enhancing the contrast, so as to obtain an RGB color space image after image preprocessing;

[0077] S3: Detect the preprocessed defect images of different types respectively to determine the defect areas, including:

[0078] S31: extracting features from the preprocessed surface image of the integrated circuit packaging process;

[0079] The surface images of integrated circuit packages collected online often introduce noise. Even if the noise reduction operation is performed, the noise cannot be completely removed. Since the low-rank embedding method can effectively remove noise and improve the robustness of the data, this embodiment uses a regularization method to optimize and improve the low-rank embedding algorithm to achieve surface defect feature extraction in the integrated circuit packaging process.

[0080] S32: performing classification operation on the image after feature extraction;

[0081] In this embodiment, according to the defect characteristics, the defect types can be divided into: manufacturing-related line width / line spacing related scale parameter defects, manufacturing-related hole related parameter defects, manufacturing-related irregularity defects, manufacturing-related unevenness defects, pollution-related color defects and trauma-related scratch defects.

[0082] Since Lie group machine learning makes full use of the advantages of manifold learning and Lie group ideas, it can describe data from a geometric perspective and provide solutions to problems from an algebraic perspective. Lie group machine learning classifiers include symplectic group classifiers and quantum group classifiers. Therefore, in this embodiment, the advantages of Lie group machine learning are combined to design a symplectic group classifier suitable for surface images of integrated circuit packaging processes, solve the difficulties faced by the classification of diverse defects in the integrated circuit packaging process, achieve accurate classification of surface defect images in the integrated circuit packaging process, and form a defect type library.

[0083] The specific steps of the symplectic group classifier algorithm include:

[0084] S321: Map the image data set to Sp(2n), and take the n-dimensional row vectors on Sp(2n) to form a set

[0085]

[0086] Among them, the vector (I 1 ,I 2 ,…,I n ,) represents any sample image.

[0087] S322: Select Q i ∈Sp(2n), convert Q into the corresponding symplectic matrix and find its singular values;

[0088] S323: Q i A sample dataset that works on images, namely:

[0089]

[0090] Formula (2) can be used as a training sample for the learning process.

[0091] S324: Perform singular value decomposition on the training sample, take the first k largest singular values, and construct a discriminant function;

[0092] S325: Solving the discriminant function can output the category of the sample.

[0093] S33: Using a corresponding defect detection method for each type of defect to effectively separate the normal background from the defect area. The specific steps include:

[0094] S331: Analyze the characteristics of various defects and the connections and characteristics between defects;

[0095] S332: For manufacturing-related line width / line spacing scale parameter defects and manufacturing-related hole parameter defects, a surface defect detection algorithm for low-density integrated circuit packaging process is used;

[0096] S333: Figure 2 As shown in the figure, for traumatic scratches and manufacturing-related uneven defects, the defect area is segmented by combining the Jardon curve related theory, including:

[0097] (1) Convert the preprocessed RGB color space image into an HSV color space image to obtain the brightness V component of the color space image;

[0098] (2) Combining the Jordan curve theorem, design a segmentation curve suitable for integrated circuit images;

[0099] In this embodiment, the Jordan curve theorem is mapped from a two-dimensional Euclidean plane to a high-dimensional high-genus surface;

[0100] In this embodiment, the Jordan curve theorem is: if C is a simple closed curve on a plane, then R 2 \C has exactly two domains, each of which has C as its boundary. Among them, the region partition can be achieved by mapping the two-dimensional Euclidean plane to a high-dimensional high-genus surface.

[0101] (3) Determine whether there are traumatic scratches and manufacturing-related uneven defects;

[0102] Since the present embodiment first performs defect classification operation on the image after feature extraction according to the defect features, and then adopts the corresponding defect detection method for each type of defect, it can be determined whether there is a corresponding defect after the region segmentation is completed.

[0103] S334: Figure 3 As shown in the figure, for pollution-type color defects, the defect area is segmented by combining the relevant theories of the Riemann manifold method, including:

[0104] (1) Collect a set number of image background samples online, design a Riemann curvature algorithm suitable for integrated circuit images, and use it to establish an image background sample fitting model;

[0105] In this embodiment, since the image background is affected by factors such as lighting, multiple pixel values ​​will appear at the same position in the Euclidean space. Therefore, the image background samples are first mapped to the Riemann manifold space. Combined with the Riemann curvature, a function is designed to calculate the optimal pixel value at the same position, thereby realizing the construction of the sample fitting model.

[0106] (2) Combined with the Riemannian metric theory, a similarity model between the image background sample and the image to be tested is constructed to calculate the similarity between the two;

[0107] In this embodiment, both the background and the image to be tested are mapped to the Riemannian manifold space, and the Riemannian metric can be used not only to calculate the length of the curve, but also to define the curvature, tangent vector, and volume of the Riemannian manifold, etc. In this embodiment, the distance between the background and the image to be tested is defined by using the Riemannian metric, and a similarity function between the two is constructed, so that the similarity between the two can be calculated.

[0108] (3) Using the principle of low-probability events to obtain the optimal threshold;

[0109] (4) The area larger than the optimal threshold is judged as background, and the area smaller than the optimal threshold is judged as defect;

[0110] (5) Realize regional segmentation of pollution-type color defects.

[0111] S335: Figure 4 As shown in the figure, for manufacturing irregular defects, defect detection is achieved by combining the relevant theory of cross-sectional curvature, including:

[0112] (1) The influence of integrated circuit texture structure on line edge extraction is removed by adaptive L*a*b color space morphological processing method;

[0113] In this embodiment, the preprocessed RGB color space image is first converted into the Lab color space, where the L component represents brightness, and a and b represent color opposition dimensions. In this embodiment, the corresponding relationship between the RGB color space and the Lab color space is designed so that the L component is not affected by the texture structure of the integrated circuit, and then the morphological method is used to further eliminate the influence of the texture structure, and finally the edge is extracted in the L component. This process can remove the influence of the texture structure of the integrated circuit on the line edge.

[0114] (2) Optimize the line edges extracted from the L component of the integrated circuit image by combining the principle of numerical approximation;

[0115] In this embodiment, since the circuit edge of the integrated circuit has been obtained, but there may be edge discontinuities during the acquisition process, it is necessary to perform a fitting operation in combination with the numerical approximation principle to optimize the edge. The edge approximation problem can be described as:

[0116] For any f(x)∈C[a,b], find an element in the subspace Φ such that and In some sense, it is the smallest. Among them, the best approximation can be described as:

[0117] Given f(x)∈C[a,b], if P *(x)∈H n Make the error:

[0118]

[0119] Now solve P * (x) is to find the maximum error max on [a,b] a≤x≤b |f(x)-P(x)| is the smallest polynomial. After solving the polynomial, the optimization of the integrated circuit circuit edge can be achieved.

[0120] (3) Combine the relevant theorems of cross-sectional curvature to calculate the constant cross-sectional curvature;

[0121] In this embodiment, the common manufacturing type irregular defect is a poor hole defect. The Riemann manifold of constant cross-sectional curvature is the simplest type, which can be divided into negative curvature, zero curvature and positive curvature. Among them, positive curvature is used to describe elliptical geometry. At this time, the approximate condition of the hole can be determined according to the value of the constant cross-sectional curvature.

[0122] (4) Determine whether there are manufacturing irregularities based on the constant cross-sectional curvature;

[0123] S34: Determine the defect area of ​​each defect type.

[0124] S4: Determine whether the defect meets industrial production requirements.

[0125] In this embodiment, the types of surface defects in the integrated circuit packaging process are diverse, and the production requirements for each type of defect are different. After the defects are segmented in step S3, it is determined whether they meet the industrial production requirements based on the shape, size and other attributes of the defect area. If they meet the industrial production requirements, they are put into subsequent production, otherwise they are discarded.

[0126] Example 2

[0127] This embodiment provides a surface defect detection system for an integrated circuit packaging process, which is used to implement the surface defect detection method for an integrated circuit packaging process of embodiment 1. The system includes: a microscopic imaging acquisition device, an image preprocessing module, a defect detection module, and a surface defect detection result output module;

[0128] In this embodiment, the microscopic imaging acquisition device is used to acquire surface digital images of the integrated circuit packaging process;

[0129] In this embodiment, the microscopic imaging acquisition device adopts a high-power metallographic microscope;

[0130] In this embodiment, the image preprocessing module is used to perform image preprocessing operations on the collected surface digital image;

[0131] In this embodiment, the defect detection module is used to detect different types of defect images after preprocessing and determine the defect area, which specifically includes:

[0132] Extract features from the pre-processed surface image of the integrated circuit packaging process;

[0133] According to the defect features, the defect type classification operation is performed on the image after feature extraction;

[0134] Use corresponding defect detection methods for each type of defect to effectively separate the normal background and defect area;

[0135] Determine the defect area for each defect type;

[0136] In this embodiment, the surface defect detection result output module is used to determine whether the defects meet the industrial production requirements and output the surface defect detection results.

[0137] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A surface defect detection method for integrated circuit packaging process, characterized in that: The steps include: Collect surface digital images of integrated circuit packaging process; Performing image preprocessing operations on the collected surface digital images; Extract features from the pre-processed surface image of the integrated circuit packaging process; According to the defect features, the defect type classification operation is performed on the image after feature extraction; Use corresponding defect detection methods for each type of defect to effectively separate the normal background and defect area; Determine the defect area for each defect type; Determine whether the defects meet industrial production requirements and output surface defect detection results.

2. The surface defect detection method for integrated circuit packaging process according to claim 1, characterized in that: The feature extraction is performed on the preprocessed surface image of the integrated circuit packaging process, and specifically a regularization method is used to optimize and improve the low-rank embedding algorithm to extract the feature of the preprocessed surface image of the integrated circuit packaging process.

3. The surface defect detection method for integrated circuit packaging process according to claim 1, characterized in that: The defect type classification operation is performed on the image after feature extraction according to the defect features, and specifically, the defect type classification operation is performed on the image after feature extraction based on a symplectic group classifier.

4. The surface defect detection method for integrated circuit packaging process according to claim 3, characterized in that: Based on the symplectic group classifier, the defect type classification operation is performed on the image after feature extraction, including: Map the image dataset to Sp(2n) and take the n-dimensional row vectors on Sp(2n) to form a set Specifically expressed as: Among them, the vector (I1,I2,…,I n ,) represents any image sample; Select Q i ∈Sp(2n), convert Q into the corresponding symplectic matrix and find its singular values; Q i Acting on the image dataset, it is expressed as: Perform singular value decomposition on the training image samples, take the first k largest singular values, and construct the discriminant function; Solve the discriminant function and output the category of the image dataset.

5. The surface defect detection method for integrated circuit packaging process according to claim 3, characterized in that: The defect types include: manufacturing-related line width or line spacing related scale parameter defects, manufacturing-related hole related parameter defects, manufacturing-related irregularity defects, manufacturing-related unevenness defects, pollution-related color defects and trauma-related scratch defects.

6. The surface defect detection method for integrated circuit packaging process according to claim 5, characterized in that: The method of using a corresponding defect detection method for each type of defect to effectively separate the normal background from the defect area specifically includes: Analyze the characteristics of various defects and the connections and characteristics between defects; Detect manufacturing-related line width or line spacing scale parameter defects, manufacturing-related hole parameter defects; Based on the Jardon curve theorem, the region segmentation of traumatic scratches and manufacturing-related non-uniform defects is performed; Regional segmentation of contamination-type color defects based on the Riemann manifold method; The line edges of the preprocessed surface image are optimized, the constant cross-sectional curvature is calculated, and whether there are manufacturing irregular defects is determined based on the constant cross-sectional curvature.

7. The surface defect detection method for integrated circuit packaging process according to claim 6, characterized in that: Based on the Jardon curve theorem, the region segmentation of traumatic scratches and manufacturing-related uneven defects is performed, including: The pre-processed surface image of the integrated circuit packaging process is converted into an HSV color space image, and a lightness V component of the color space image is obtained; Construct the segmentation curve of surface image based on Jordan curve theorem; Regional segmentation of trauma-related scratches and manufacturing-related uneven defects based on segmentation curves; Based on the segmented area, determine whether there are trauma-related scratches and manufacturing-related uneven defects.

8. The surface defect detection method for integrated circuit packaging process according to claim 6, characterized in that: The regional segmentation of pollution-type color defects is performed based on the Riemann manifold method, specifically including: Obtain image background samples, and map the image background samples to the Riemann manifold space; Constructing a similarity model between the image background sample and the image to be tested, and calculating the similarity between the image background sample and the image to be tested; Use the principle of low-probability events to obtain the optimal threshold; The area larger than the optimal threshold is judged as the background, and the area smaller than the optimal threshold is judged as a defect, thus completing the regional segmentation of contamination-type color defects.

9. The surface defect detection method for integrated circuit packaging process according to claim 6, characterized in that: Optimize the line edges of the preprocessed surface image, including: The preprocessed surface image is converted from RGB color space to Lab color space, and image processing is performed based on morphological methods to extract line edges in the L component of the Lab color space; A fitting operation is performed based on the principle of numerical approximation to optimize the line edges extracted from the L component.

10. A surface defect detection system for integrated circuit packaging process, characterized in that: For implementing the surface defect detection method for integrated circuit packaging process of claims 1-9, the system comprises: a microscopic imaging acquisition device, an image preprocessing module, a defect detection module, and a surface defect detection result output module; The microscopic imaging acquisition device is used to acquire surface digital images of the integrated circuit packaging process; The image preprocessing module is used to perform image preprocessing operations on the collected surface digital images; The defect detection module is used to detect different types of defect images after preprocessing and determine the defect area, specifically including: Extract features from the pre-processed surface image of the integrated circuit packaging process; According to the defect features, the defect type classification operation is performed on the image after feature extraction; Use corresponding defect detection methods for each type of defect to effectively separate the normal background and defect area; Determine the defect area for each defect type; The surface defect detection result output module is used to determine whether the defects meet the industrial production requirements and output the surface defect detection results.

Images