Wafer bonding defect detection method and device, computer equipment and storage medium
By combining instance segmentation models and visual inspection, and optimizing boundary segmentation and feature extraction, the problem of difficulty in identifying minute defects in traditional detection techniques is solved, and efficient and high-precision wafer bonding interface defect detection is achieved.
Patent Information
- Application Number
- CN202510974660.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-15
- Publication Date
- 2025-11-28
AI Technical Summary
Traditional inspection techniques struggle to accurately identify minute defects at wafer bonding interfaces in complex environments, deep learning models are susceptible to noise interference, and visual inspection lacks robustness.
An instance segmentation model is used for initial segmentation. Visual inspection is then used to correct missegmented regions and optimize boundaries. Geometric and morphological features of defective objects are extracted for defect identification.
This improved the accuracy and efficiency of defect detection at wafer bonding interfaces, achieving high-precision defect identification.
Smart Images
Figure CN121032899A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of wafer inspection technology, and in particular to a method, apparatus, computer equipment and storage medium for detecting wafer bonding defects. Background Technology
[0002] As semiconductor processes evolve towards 3D integration and advanced packaging, defects (voids, cracks, delamination) at wafer bonding interfaces (such as copper-copper direct bonding and hybrid bonding) have become a key bottleneck affecting chip yield.
[0003] In traditional detection techniques, relying solely on deep learning models is susceptible to interference from complex backgrounds (such as bonding interface textures and noise), resulting in a high missegmentation rate. While traditional visual detection techniques (such as edge detection) are robust, their ability to identify minute defects is limited. Summary of the Invention
[0004] This application provides a wafer bonding defect detection method, apparatus, computer equipment, and storage medium, which can perform high-accuracy and high-efficiency defect detection on wafer bonding interfaces.
[0005] According to one aspect of the embodiments of this application, a wafer bonding defect detection method is provided, comprising: Acquire raw images of the wafer bonding interface; The original acquired image is input into the instance segmentation model for segmentation processing to identify defect objects at the wafer bonding interface and generate a boundary segmentation mask; The boundary segmentation mask is further verified to correct mis-segmented regions and optimize boundary accuracy, resulting in an optimized segmentation result. Based on the optimized segmentation results, the geometric and morphological features of the defective object are extracted. Based on the geometric and morphological feature information, defect discrimination processing is performed, and a defect detection report of the wafer bonding interface is output.
[0006] Optionally, acquiring the raw image of the wafer bonding interface includes: The image acquisition device is controlled to emit ultrasonic signals to the wafer bonding interface; Receive the reflected signal reflected from the wafer bonding interface; The original acquired image is generated based on the reflected signal.
[0007] Optionally, the step of inputting the original acquired image into an instance segmentation model for segmentation processing, identifying defect objects at the wafer bonding interface, and generating a boundary segmentation mask includes: The original acquired image is input into the instance segmentation model for segmentation processing to identify defect objects appearing in the wafer bonding interface; Determine the pixel coordinate data corresponding to the defective object; Generate a boundary segmentation mask based on the pixel coordinate data.
[0008] Optionally, the step of supplementing and verifying the boundary segmentation mask, correcting mis-segmented regions, and optimizing boundary accuracy to obtain an optimized segmentation result includes: The boundary segmentation mask is further verified to obtain the misjudgment result corresponding to the defective object; If the misjudgment result indicates that the first defective object belongs to the normal area, the segmented area corresponding to the first defective object is determined as the missegmented area. Boundary optimization processing is performed on the segmented region corresponding to the second defective object to obtain the optimized segmented region corresponding to the second defective object; The boundary segmentation mask is updated based on the missegmented region and the optimized segmented region to obtain the optimized segmentation result.
[0009] Optionally, the step of extracting the geometric and morphological feature information of the defective object based on the optimized segmentation results includes: If the optimized segmentation result includes at least one of the second defect objects, then edge detection processing is performed on the optimized segmentation region corresponding to at least one of the second defect objects to obtain geometric feature information corresponding to at least one of the second defect objects, and feature matching processing is performed on the optimized segmentation region corresponding to at least one of the second defect objects to obtain morphological feature information corresponding to at least one of the second defect objects.
[0010] Optionally, the geometric feature information includes defect size information, defect area information, and defect location information; the morphological feature information includes contour complexity information and grayscale distribution information.
[0011] Optionally, the defect discrimination processing based on the geometric feature information and the morphological feature information, and the output of a defect detection report for the wafer bonding interface, includes: The defect size information, defect area information, defect location information, contour complexity information, and grayscale distribution information corresponding to at least one of the second defect objects are output to the defect classification model, and the probability value of each second defect object corresponding to each of the multiple defect types is output; the multiple defect types include void defects, crack defects, delamination defects, and cold weld defects. Determine the maximum probability value corresponding to each second defect object, and determine the defect type corresponding to the maximum probability value as the identification defect type corresponding to the second defect object; The defect detection report is generated based on the identified defect type corresponding to each of the second defect objects.
[0012] According to one aspect of the embodiments of this application, a wafer bonding defect detection device is provided, characterized in that the device comprises: The image acquisition module is used to acquire raw images of the wafer bonding interface; The image segmentation module is used to input the original acquired image into the instance segmentation model for segmentation processing, identify the defect objects of the wafer bonding interface and generate a boundary segmentation mask; The image verification module is used to perform supplementary verification on the boundary segmentation mask, correct mis-segmented regions and optimize boundary accuracy to obtain optimized segmentation results; The defect identification module is used to extract the geometric and morphological feature information of the defective object based on the optimized segmentation results. The defect classification module is used to perform defect discrimination processing based on the geometric feature information and the morphological feature information, and output a defect detection report of the wafer bonding interface.
[0013] According to one aspect of the embodiments of this application, a computer device is provided, the computer device including a processor and a memory, the memory storing at least one instruction, at least one program, code set or instruction set, the at least one instruction, the at least one program, the code set or instruction set being loaded and executed by the processor to implement the above-described wafer bonding defect detection method.
[0014] According to one aspect of the embodiments of this application, a computer-readable storage medium is provided, wherein at least one instruction, at least one program, code set, or instruction set is stored in the storage medium, and the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the above-described wafer bonding defect detection method.
[0015] According to one aspect of the embodiments of this application, a computer program product is provided, the computer program product including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, causing the computer device to perform to implement the above-described wafer bonding defect detection method.
[0016] The technical solution provided in this application can bring the following beneficial effects: The method involves acquiring an initial image of the wafer bonding interface; inputting the initial image into an instance segmentation model for segmentation processing to identify defect objects on the wafer bonding interface and generate a boundary segmentation mask; further verifying the boundary segmentation mask to correct mis-segmented areas and optimize boundary accuracy, resulting in an optimized segmentation result; extracting the geometric and morphological features of the defect objects based on the optimized segmentation result; performing defect discrimination processing based on the geometric and morphological features, and outputting a defect detection report for the wafer bonding interface. In this method, high-precision initial segmentation is achieved first through a deep learning-based instance segmentation model, followed by visual detection processing to correct mis-segmented areas and optimize boundaries, and then defect discrimination is performed. This approach improves the detection accuracy and efficiency of defect detection on wafer bonding interfaces. Attached Figure Description
[0017] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 This is a schematic flowchart of a wafer bonding defect detection method provided in one embodiment of this application. Detailed Implementation
[0019] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.
[0020] Please refer to Figure 1 This document illustrates a flowchart of a wafer bonding defect detection method according to an embodiment of this application. The method can be applied to computer equipment, which refers to electronic devices capable of data computation and processing. The method may include the following steps (110-150): Step 110: Acquire the raw image of the wafer bonding interface.
[0021] Step 120: Input the original acquired image into the instance segmentation model for segmentation processing, identify defect objects at the wafer bonding interface and generate boundary segmentation masks.
[0022] Step 130: Perform supplementary verification on the boundary segmentation mask, correct mis-segmented regions and optimize boundary accuracy to obtain the optimized segmentation result.
[0023] Step 140: Based on the optimized segmentation results, extract the geometric and morphological features of the defective object.
[0024] Step 150: Perform defect discrimination processing based on geometric and morphological feature information, and output a defect detection report of the wafer bonding interface.
[0025] Optionally, step 110 above specifically includes the following processes: The image acquisition device is controlled to emit ultrasonic signals to the wafer bonding interface; Receive the reflected signal from the wafer bonding interface; The original acquired image is generated based on the reflected signal.
[0026] Optionally, step 120 above specifically includes the following processes: The original acquired images are input into the instance segmentation model for segmentation processing to identify defective objects appearing in the wafer bonding interface. Determine the pixel coordinate data corresponding to the defective object; Generate a boundary segmentation mask based on pixel coordinate data.
[0027] Optionally, step 130 above specifically includes the following processes: Supplementary verification processing is performed on the boundary segmentation mask to obtain the misjudgment result corresponding to the defective object; If the misjudgment result indicates that the first defective object belongs to the normal area, the segmentation area corresponding to the first defective object shall be determined as the missegmentation area. Boundary optimization processing is performed on the segmented region corresponding to the second defective object to obtain the optimized segmented region corresponding to the second defective object; The boundary segmentation mask is updated based on the missegmented region and the optimized segmented region to obtain the optimized segmentation result.
[0028] Optionally, step 140 above specifically includes the following processes: If the optimized segmentation result includes at least one second defect object, then edge detection processing is performed on the optimized segmentation region corresponding to at least one second defect object to obtain the geometric feature information corresponding to at least one second defect object, and feature matching processing is performed on the optimized segmentation region corresponding to at least one second defect object to obtain the morphological feature information corresponding to at least one second defect object.
[0029] Optionally, the geometric feature information includes defect size information, defect area information, and defect location information; the morphological feature information includes contour complexity information and grayscale distribution information.
[0030] Optionally, step 150 above specifically includes the following processes: The defect size information, defect area information, defect location information, contour complexity information and grayscale distribution information corresponding to at least one second defect object are output to the defect classification model. The probability value of each second defect object corresponding to each of the multiple defect types is output. The multiple defect types include void defects, crack defects, delamination defects and cold weld defects. Determine the maximum probability value corresponding to each second defect object, and determine the defect type corresponding to the maximum probability value as the identification defect type corresponding to the second defect object; A defect detection report is generated based on the identified defect type corresponding to each second defect object.
[0031] In summary, the technical solution provided in this application involves: acquiring the original image of the wafer bonding interface; inputting the original image into an instance segmentation model for segmentation processing to identify defect objects on the wafer bonding interface and generate a boundary segmentation mask; further verifying the boundary segmentation mask to correct mis-segmented areas and optimize boundary accuracy, resulting in an optimized segmentation result; extracting the geometric and morphological feature information of the defect object based on the optimized segmentation result; performing defect discrimination processing based on the geometric and morphological feature information, and outputting a defect detection report for the wafer bonding interface. In the technical solution provided in this application, high-precision preliminary segmentation is first achieved through a deep learning-based instance segmentation model, followed by visual detection processing to correct mis-segmented areas and optimize boundaries, and then defect discrimination is performed. This approach improves the detection accuracy and efficiency of defect detection on the wafer bonding interface. The following are embodiments of the apparatus of this application, which can be used to execute embodiments of the method of this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the method of this application.
[0032] One embodiment of this application provides a wafer bonding defect detection device. This device has the function of implementing the above-described wafer bonding defect detection method. The function can be implemented in hardware or by hardware executing corresponding software. The device can be a computer device or can be installed within a computer device. The wafer bonding defect detection device may include: The image segmentation module is used to input the original acquired image into the instance segmentation model for segmentation processing, identify defect objects at the wafer bonding interface, and generate boundary segmentation masks. The image verification module is used to supplement and verify the boundary segmentation mask, correct mis-segmented regions and optimize boundary accuracy to obtain the optimized segmentation result. The defect identification module is used to extract the geometric and morphological features of defective objects based on the optimized segmentation results. The defect classification module is used to perform defect discrimination processing based on geometric and morphological feature information, and output a defect detection report of the wafer bonding interface.
[0033] It should be noted that the apparatus provided in the above embodiments is only illustrated by the division of the above functional modules when implementing its functions. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the apparatus and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here. One embodiment of this application provides a computer device for implementing the wafer bonding defect detection method provided in the above embodiments. Specifically: Typically, computer equipment includes a processor and memory.
[0034] The processor may include one or more processing cores, such as a quad-core processor or an octa-core processor. The processor may be implemented using at least one hardware form of DSP (Digital Signal Processing), FPGA (Field Programmable Gate Array), or PLA (Programmable Logic Array). The processor may also include a main processor and coprocessors. The main processor, also known as the CPU (Central Processing Unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, the processor may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, the processor may also include an AI (Artificial Intelligence) processor, which handles computational operations related to machine learning.
[0035] The memory may include one or more computer-readable storage media, which may be non-transitory. The memory may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments, the non-transitory computer-readable storage media in the memory is used to store at least one instruction, at least one program, code set, or instruction set, configured to be executed by one or more processors to implement the wafer bonding defect detection method described above.
[0036] In some embodiments, the computer device may also optionally include: a peripheral device interface and at least one peripheral device. The processor, memory, and peripheral device interface can be connected via a bus or signal line. Each peripheral device can be connected to the peripheral device interface via a bus, signal line, or circuit board. Specifically, the peripheral device includes at least one of: a power source drive module, a touch display screen, a temperature control component, an audio circuit, and a power supply.
[0037] Computer equipment can receive user input to execute the steps in the above method or the operations within those steps.
[0038] Those skilled in the art will understand that the structure described above does not constitute a limitation on the computer device, and may include more or fewer components, or combine certain components, or use different component arrangements.
[0039] In an exemplary embodiment, a computer-readable storage medium is also provided, the storage medium storing at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set, when executed by a processor, implements the above-described wafer bonding defect detection method.
[0040] Optionally, the computer-readable storage medium may include: ROM (Read Only Memory), RAM (Random Access Memory), SSD (Solid State Drives), or optical disc, etc. The random access memory may include ReRAM (Resistance Random Access Memory) and DRAM (Dynamic Random Access Memory).
[0041] In an exemplary embodiment, a computer program product or computer program is also provided, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the aforementioned wafer bonding defect detection method.
[0042] It should be understood that "multiple" as used herein refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. Furthermore, the step numbers described herein are merely illustrative of one possible execution order. In some other embodiments, the steps may not be executed in numerical order, such as two steps with different numbers being executed simultaneously, or two steps with different numbers being executed in the reverse order of the illustration. This application does not limit this.
[0043] In the description of this application, it should be noted that, in the embodiments of this application, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature.
[0044] In the description of the embodiments of this application, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, "linking" can be a detachable connection or a non-detachable connection; it can be a direct connection or an indirect connection through an intermediate medium. "Fixed connection" refers to a connection where the relative positional relationship remains unchanged after the connection.
[0045] The directional terms used in the embodiments of this application, such as "inner" and "outer," are merely for reference to the directions in the accompanying drawings. Therefore, the directional terms used are for better and clearer explanation and understanding of the embodiments of this application, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of this application. Furthermore, unless otherwise stated in this application, "multiple" as used in this application refers to two or more.
[0046] In the description of embodiments of this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0047] The above description is merely an exemplary embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.
Claims
1. A method for detecting wafer bonding defects, characterized in that, The method includes: Acquire raw images of the wafer bonding interface; The original acquired image is input into the instance segmentation model for segmentation processing to identify defect objects at the wafer bonding interface and generate a boundary segmentation mask; The boundary segmentation mask is further verified to correct mis-segmented regions and optimize boundary accuracy, resulting in an optimized segmentation result. Based on the optimized segmentation results, the geometric and morphological features of the defective object are extracted. Based on the geometric and morphological feature information, defect discrimination processing is performed, and a defect detection report of the wafer bonding interface is output.
2. The method according to claim 1, characterized in that, The acquisition of the original image of the wafer bonding interface includes: The image acquisition device is controlled to emit ultrasonic signals to the wafer bonding interface; Receive the reflected signal reflected from the wafer bonding interface; The original acquired image is generated based on the reflected signal.
3. The method according to claim 1, characterized in that, The step of inputting the original acquired image into the instance segmentation model for segmentation processing, identifying defect objects at the wafer bonding interface, and generating a boundary segmentation mask includes: The original acquired image is input into the instance segmentation model for segmentation processing to identify defect objects appearing in the wafer bonding interface; Determine the pixel coordinate data corresponding to the defective object; Generate a boundary segmentation mask based on the pixel coordinate data.
4. The method according to claim 1, characterized in that, The step of supplementing and verifying the boundary segmentation mask, correcting mis-segmented regions, and optimizing boundary accuracy to obtain an optimized segmentation result includes: The boundary segmentation mask is further verified to obtain the misjudgment result corresponding to the defective object; If the misjudgment result indicates that the first defective object belongs to the normal area, the segmented area corresponding to the first defective object is determined as the missegmented area. Boundary optimization processing is performed on the segmented region corresponding to the second defective object to obtain the optimized segmented region corresponding to the second defective object; The boundary segmentation mask is updated based on the missegmented region and the optimized segmented region to obtain the optimized segmentation result.
5. The method according to claim 4, characterized in that, Based on the optimized segmentation results, the geometric and morphological feature information of the defective object is extracted, including: If the optimized segmentation result includes at least one of the second defect objects, then edge detection processing is performed on the optimized segmentation region corresponding to at least one of the second defect objects to obtain geometric feature information corresponding to at least one of the second defect objects, and feature matching processing is performed on the optimized segmentation region corresponding to at least one of the second defect objects to obtain morphological feature information corresponding to at least one of the second defect objects.
6. The method according to claim 5, characterized in that, The geometric feature information includes defect size information, defect area information, and defect location information; the morphological feature information includes contour complexity information and grayscale distribution information.
7. The method according to claim 6, characterized in that, The defect discrimination processing based on the geometric feature information and the morphological feature information, and the output of the defect detection report of the wafer bonding interface, includes: The defect size information, defect area information, defect location information, contour complexity information, and grayscale distribution information corresponding to at least one of the second defect objects are output to the defect classification model, and the probability value of each second defect object corresponding to each of the multiple defect types is output; the multiple defect types include void defects, crack defects, delamination defects, and cold weld defects. Determine the maximum probability value corresponding to each second defect object, and determine the defect type corresponding to the maximum probability value as the identification defect type corresponding to the second defect object; The defect detection report is generated based on the identified defect type corresponding to each of the second defect objects.
8. A wafer bonding defect detection device, characterized in that, The device includes: The image acquisition module is used to acquire raw images of the wafer bonding interface; The image segmentation module is used to input the original acquired image into the instance segmentation model for segmentation processing, identify the defect objects of the wafer bonding interface and generate a boundary segmentation mask; The image verification module is used to perform supplementary verification on the boundary segmentation mask, correct mis-segmented regions and optimize boundary accuracy to obtain optimized segmentation results; The defect identification module is used to extract the geometric and morphological feature information of the defective object based on the optimized segmentation results. The defect classification module is used to perform defect discrimination processing based on the geometric feature information and the morphological feature information, and output a defect detection report of the wafer bonding interface.
9. A computer device comprising a processor and a memory, the memory storing at least one instruction, at least one program, a code set, or an instruction set, the at least one instruction, the at least one program, the code set, or the instruction set being loaded and executed by the processor to implement the method as claimed in any one of claims 1 to 7.
10. A computer-readable storage medium storing at least one instruction, at least one program, a code set, or an instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the method as claimed in any one of claims 1 to 7.