Image Processing Method, Apparatus, Device, and Storage Medium

Abstract

The present application relates to an image processing method, apparatus, device and storage medium, which are applied in the field of wafer defect detection. It includes applying to any starting position on the back of the collected wafer, determining the starting focal length sequence of the microscope lens according to the minimum focusing distance of the obtained microscope lens and the allowable height difference of the wafer; obtaining the starting image sequence collected by the microscope lens at the starting position, wherein the starting image sequence is adapted to the starting focal length sequence; calculating the clarity of each image in the starting image sequence based on a preset image clarity evaluation algorithm, determining an image subsequence according to a preset sequence length, and determining the image subsequence with the highest clarity score as the starting target image subsequence; performing image fusion processing on the images in the starting target image subsequence to obtain the target clear image at the starting position. The technical effects of the present application are: reducing the acquisition time of wafer images and the image fusion processing time of wafer images, and improving the efficiency of wafer defect detection.

Description

Technical Field

[0001] The present application relates to the technical field of wafer defect detection, and in particular, to an image processing method, apparatus, device, and storage medium. Background Art

[0002] In the process of semiconductor wafer manufacturing, the processing quality of the wafer directly determines the yield and quality of subsequent chip products. The surface of the wafer contains circuits of the chip, etc. Affected by various factors such as process and raw materials, the defects on the wafer surface are usually very tiny and have diverse morphological structures. Therefore, multiple high-precision and clear wafer images usually need to be obtained to detect the defects on the wafer surface. At the same time, during the wafer manufacturing process, there may also be defects such as scratches and corrosion on the back surface of the wafer. The scratches, corrosion, etc. on the back surface may cause cracks in the chip. Therefore, the defects on the back surface of the wafer are detected to ensure the wafer quality and improve the wafer yield.

[0003] The defect detection of the wafer back surface is carried out according to the collected wafer images. Due to the non-absolutely flatness of the wafer back surface, that is, there is a height difference on the wafer, in order to obtain a clear image, it is necessary to take pictures of the wafer at different focal lengths and perform image fusion to obtain a clear image and improve the image quality, thereby improving the defect detection rate; however, the structure of the wafer back surface is relatively simple. Performing image fusion on all the collected images will result in a large amount of image data to be processed and a long image processing time, thereby reducing the efficiency of wafer defect detection. Summary of the Invention

[0004] In order to help solve the problem that the structure of the wafer back surface is relatively simple, performing image fusion on all the collected images will result in a large amount of image data to be processed, a long image processing time, and thus a reduction in the efficiency of wafer defect detection, the present application provides an image processing method, apparatus, device, and storage medium.

[0005] In a first aspect, the present application provides an image processing method, adopting the following technical solution: The method is applied to an automatic optical inspection system for wafer back surface defects. The automatic optical inspection system for wafer back surface defects includes a microscopic lens, and the microscopic lens is used to collect images of the wafer back surface. The focal length of the microscopic lens is adjustable. The method includes:

[0006] Applied to any starting position of the wafer back surface to be collected, determine the starting focal length sequence of the microscopic lens according to the obtained minimum focusing distance of the microscopic lens and the allowable height difference of the wafer;

[0007] Obtain the starting image sequence collected by the microscopic lens at the starting position, where the starting image sequence is adapted to the starting focal length sequence;

[0008] Calculate the clarity of each image in the starting image sequence based on a preset image clarity evaluation algorithm, determine an image subsequence according to a preset sequence length, and determine the image subsequence with the highest clarity score as the starting target image subsequence;

[0009] Perform image fusion processing on the images in the starting target image subsequence to obtain the target clear image at the starting position.

[0010] In a specific feasible implementation, after performing image fusion processing on the images in the starting target image subsequence to obtain the target clear image at the starting position, it further includes:

[0011] Applied to any non-starting position on the back of the collected wafer, determine the target focal length at the current position according to the target image subsequence at the previous position;

[0012] Obtain the first image collected by the microscope lens at the current position, where the first image is adapted to the target focal length;

[0013] Calculate the clarity of the first image based on a preset image clarity evaluation algorithm, determine whether the clarity of the first image is within a first preset threshold range, and determine the target clear image at the current position according to the judgment result.

[0014] In a specific feasible implementation, determining the target clear image at the current position according to the judgment result includes:

[0015] If the clarity value of the first image is within the first preset threshold range, determine that the first image is the target clear image at the current position;

[0016] If the clarity value of the first image is not within the first preset threshold range, determine an initial focal length subsequence at the current position according to the target focal length at the current position, and determine the target clear image at the current position according to the initial focal length subsequence at the current position.

[0017] In a specific feasible implementation, determining the target clear image at the current position according to the initial focal length subsequence at the current position includes:

[0018] Obtain the initial image subsequence collected by the microscope lens at the current position, where the initial image subsequence is adapted to the initial focal length subsequence at the current position;

[0019] Calculate the sharpness of each image in the initial image subsequence based on a preset image sharpness evaluation algorithm. If the sharpness value of each image is within the second preset threshold range, determine that the initial image subsequence is the target image subsequence at the current position;

[0020] Perform image fusion processing on the images in the target image subsequence to obtain the target clear image at the current position.

[0021] In a specific feasible implementation, before performing image fusion processing on the images in the target image subsequence to obtain the target clear image at the current position, it further includes:

[0022] If the sharpness value of each image is not within the second preset threshold range, adjust the initial focal length subsequence at the current position according to the image sharpness calculation result and the minimum focusing distance of the microscope lens to generate the adjusted focal length subsequence at the current position;

[0023] Obtain the adjusted image subsequence corresponding to the adjusted focal length subsequence at the current position, and re-determine whether the sharpness values of the images in the adjusted image subsequence are all within the second preset threshold range. If the sharpness values of the images in the adjusted image subsequence are all within the second preset threshold range, determine that the adjusted image subsequence is the target image subsequence at the current position.

[0024] In a specific feasible implementation, determining the target focal length at the current position based on the target image subsequence at the previous position includes:

[0025] Calculate the sharpness of each image in the target image subsequence at the previous position according to a preset image sharpness evaluation algorithm, and determine the focal length corresponding to the image with the highest sharpness;

[0026] Calculate the height difference between the current position and the previous position, and determine the target focal length at the current position according to the height difference and the focal length corresponding to the image with the highest sharpness at the previous position.

[0027] In a specific feasible implementation, the preset sequence length is at least 2.

[0028] In a second aspect, the present application provides an image processing device, adopting the following technical solution: The device is applied to an automatic optical inspection system for wafer backside defects. The automatic optical inspection system for wafer backside defects includes a microscope lens for collecting images of the wafer backside, and the focal length of the microscope lens is adjustable. The device includes:

[0029] The starting sequence determination module is used to be applied to any starting position on the back of the collected wafer, and determine the starting focal length sequence of the microscope lens according to the obtained minimum focusing distance of the microscope lens and the allowable acceptance height difference of the wafer;

[0030] The starting image acquisition module is used to acquire the starting image sequence collected by the microscope lens at the starting position, where the starting image sequence is adapted to the starting focal length sequence;

[0031] The target image determination module is used to calculate the clarity of each image in the starting image sequence based on a preset image clarity evaluation algorithm, determine an image subsequence according to a preset sequence length, and determine the image subsequence with the highest clarity score as the starting target image subsequence;

[0032] The clear image generation module is used to perform image fusion processing on the images in the starting target image subsequence to obtain the target clear image at the starting position.

[0033] In a third aspect, the present application provides a computer device, adopting the following technical solution: including a memory and a processor, and a computer program capable of being loaded and executed by the processor and being the above-mentioned any image processing method is stored on the memory.

[0034] In a fourth aspect, the present application provides a computer-readable storage medium, adopting the following technical solution: storing a computer program capable of being loaded and executed by the processor and being the above-mentioned any image processing method.

[0035] In summary, the present application has the following beneficial technical effects:

[0036] By obtaining the maximum number of images on the back of the wafer and then determining the starting target image subsequence with higher clarity according to the image clarity, the number of the starting target image subsequences can be determined by the user, thereby reducing the number of image fusions, reducing the acquisition time of the wafer detection images, and at the same time saving the processing time of wafer image fusion, and further improving the efficiency of wafer image fusion and wafer defect detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is a flowchart of the image processing method in an embodiment of the present application;

[0038] Figure 2 is a schematic diagram of the image processing device in an embodiment of the present application;

[0039] Figure 3 is a schematic diagram for embodying the computer device in an embodiment of the present application.

[0040] Reference numerals: 201, starting sequence determination module; 202, starting image acquisition module; 203, target image determination module; 204, clear image generation module. DETAILED DESCRIPTION

[0041] The following combination Figures 1-3 This application is described in further detail.

[0042] The embodiment of the present application discloses an image processing method, which is applied to an automatic optical detection system for defects on the back of a wafer. The image processing method can reduce the number of images during image fusion, thereby saving the processing time of wafer image fusion, and can improve the quality of wafer images after fusion and the efficiency of wafer defect detection. In the semiconductor wafer manufacturing process, the processing quality of the wafer directly determines the yield and quality of subsequent chip products. The wafer surface contains the circuits of the chip, etc. Affected by various factors such as process and raw materials, the surface defects of the wafer are usually very small and have diverse morphological structures. Therefore, the surface defects of the wafer usually require the acquisition of multiple high-precision and clear wafer images to detect the defects on the surface of the wafer. At the same time, during the wafer manufacturing process, there may also be defects such as scratches and corrosion on the back of the wafer. The scratches and corrosion on the back may cause cracks in the chip. Therefore, the defects on the back of the wafer are detected to ensure the quality of the wafer and improve the wafer yield.

[0043] The defect detection on the back of the wafer is performed based on the collected wafer image. Due to the non-absolute flatness of the back of the wafer, that is, the height difference of the wafer, in order to obtain a clear image, it is necessary to shoot the wafer at different focal lengths and perform image fusion to obtain a clear image and improve the image quality, thereby improving the defect detection rate; however, the structure of the back of the wafer is relatively simple, and image fusion of all collected images will result in a large amount of processed image data and a long image processing time, which will lead to a decrease in the efficiency of wafer defect detection. In order to help improve the clarity of wafer fusion images and the efficiency of wafer defect detection, the present application provides an image processing method.

[0044] Reference Figure 1 , the method comprises the following steps:

[0045] S10, applied to any starting position on the back side of the wafer to be collected, determines the starting focal length sequence of the microscope lens according to the acquired minimum focusing distance of the microscope lens and the acceptable height difference of the wafer.

[0046] Specifically, the wafer image processing method is applied to an automatic optical inspection system for wafer backside defects. The image processing method is implemented based on a liquid microscope lens, which has a minimum focusing distance that can be preset by the user. At any starting position on the backside of the wafer to be collected, the system determines the starting focal length sequence of the microscope lens according to the obtained minimum focusing distance of the microscope lens and the allowable height difference of the wafer.

[0047] It should be noted that the number of images of the wafer backside obtained according to the minimum focusing distance of the microscope lens is the maximum number of images. In addition, the number of obtained wafer images is related to the allowable height difference of the wafer. The allowable height difference of the wafer is related to the warpage amplitude of the wafer. During the manufacturing process of the wafer, the wafer may have a certain warpage amplitude. Generally, the warpage amplitude of the wafer is ±5 mm, then the allowable height difference of the wafer is 10 mm. At this time, the number of the starting focal length sequence of the microscope lens is 10 mm divided by the minimum focusing distance of the microscope lens. Assuming that the minimum focusing distance of the microscope lens is 0.4 mm, then there are 25 consecutive focal lengths in the finally determined starting focal length sequence.

[0048] S20. Obtain the starting image sequence collected by the microscope lens at the starting position, where the starting image sequence is adapted to the starting focal length sequence.

[0049] Specifically, obtain the corresponding starting image sequence according to the determined starting focal length sequence. For example, assuming that the minimum focusing distance of the microscope lens is 0.4 mm and the warpage amplitude of the wafer is ±5 mm, then the allowable height difference of the wafer is 10 mm, and there are 25 consecutive focal lengths in the finally determined starting focal length sequence. The microscope lens obtains the corresponding starting image sequence according to the determined 25 focal length sequences.

[0050] S30. Calculate the clarity of each image in the starting image sequence based on a preset image clarity evaluation algorithm, determine the image subsequence according to the preset sequence length, and determine the image subsequence with the highest clarity score as the starting target image subsequence.

[0051] Specifically, the sharpness of each image in the starting image sequence is calculated based on a preset image sharpness evaluation algorithm. Among them, the image sharpness evaluation algorithm can directly use existing algorithms to evaluate the sharpness of each image and give corresponding scores. Then, an image subsequence is determined according to the preset sequence length. For example, if the length of the image sequence is set to 5, then 5 consecutive images need to be selected in the image subsequence. The selection method can be to sort the image sequence according to sharpness, and determine the image subsequence with the highest sharpness score as the starting target image subsequence. Considering that there are blurred and clear images in the maximum number of starting image sequences collected, and in the process of image fusion, blurred images cannot improve the sharpness of the fused image, and may increase the time of image fusion processing, thereby reducing the efficiency of image fusion. Therefore, by sorting according to the level of sharpness, the starting target image subsequence is determined, reducing the number of images, thus saving the processing time of image fusion, improving the efficiency of image fusion, and further improving the efficiency of wafer detection.

[0052] It should be noted that in the embodiment of the present application, the preset sequence length, that is, the number of starting target image subsequences, is at least 2. In the manufacturing process of wafers, wafers are not absolutely flat and may have a certain warping amplitude. Considering the warping amplitude of wafers, the number of consecutive clear images is determined to be at least 2 to ensure that clear images can be captured at sloped detection points for defect detection, thereby improving the defect detection rate.

[0053] S40, perform image fusion processing on the images in the starting target image subsequence to obtain the target clear image at the starting position.

[0054] Specifically, fuse the images in the determined starting target image subsequence to generate the final target clear image at the starting position. Considering that there are also clear images and blurred images in the images in the determined starting target image subsequence, through the method of image fusion, the clear parts of multiple images are fused, thereby generating a wafer fusion image with higher sharpness to obtain the final target clear image, improving the quality of the image, which helps to perform wafer defect detection based on the clear image subsequently.

[0055] In the solution of the present application, by obtaining the maximum number of backside images of wafers and then determining the starting target image subsequence with higher sharpness according to image sharpness, the number of the starting target image subsequence can be determined by the user, thus reducing the number of image fusions, reducing the acquisition time of wafer detection images, saving the processing time of wafer image fusion at the same time, and further improving the efficiency of wafer image fusion and wafer defect detection.

[0056] In one embodiment, after performing image fusion processing on the images in the starting target image subsequence to obtain the target clear image at the starting position, the following steps may further be executed:

[0057] The backside defect detection of the wafer is a continuous process. After determining the starting position of the backside defect detection of the wafer, there are several detection points at non-starting positions to be detected subsequently. Therefore, for any non-starting position on the backside of the collected wafer, the target focal length at the current position is determined according to the target image subsequence at the previous position. Specifically, the sharpness of each image in the target image subsequence at the previous position is calculated according to a preset image sharpness evaluation algorithm, and the focal length corresponding to the image with the highest sharpness is determined; the height difference between the current position and the previous position is calculated, and the target focal length at the current position is determined according to the height difference and the focal length corresponding to the image with the highest sharpness at the previous position. For example, assuming that the current position is a non-starting position immediately following the starting position, the previous position of the current position can be understood as the starting position, and then the target focal length at the current position is determined according to the starting target image subsequence obtained at the starting position; the determination method can be to first calculate the sharpness of each image in the starting target image subsequence through the sharpness evaluation algorithm, and determine the focal length corresponding to the image with the highest sharpness, and then calculate the height difference between the current position and the starting position, and on the basis of the focal length corresponding to the image with the highest sharpness determined, adjust the focal length according to the height difference to obtain the target focal length at the current position.

[0058] After determining the target focal length at the current position, a first image collected by the microscope lens at the current position is obtained according to the determined target focal length. Finally, the sharpness of the first image is calculated based on a preset image sharpness evaluation algorithm, it is judged whether the sharpness of the first image is within a first preset threshold range, and the target clear image at the current position is determined according to the judgment result.

[0059] If the sharpness value of the first image is within the first preset threshold range, it is determined that the first image is the target clear image at the current position; if the sharpness value of the first image is not within the first preset threshold range, the initial focal length subsequence at the current position is determined according to the target focal length at the current position, and the target clear image at the current position is determined according to the initial focal length subsequence at the current position. Specifically, it is judged whether the sharpness of the obtained first image is within the preset threshold range according to the sharpness evaluation algorithm. If it is within the conforming threshold range, it means that the first image is a clear image, and the obtained first image can be directly set as the target clear image. If the image sharpness conforms to the threshold, there is no need to obtain more images for image fusion, thus the time for image fusion can be saved and the efficiency of image fusion can be improved; if it is not within the conforming threshold range, an image sequence with a preset sequence length needs to be obtained according to the target focal length at the current position for image fusion to obtain a wafer backside image with higher sharpness.

[0060] When the clarity of the first image is not within the threshold range, the step of obtaining an image sequence of a preset sequence length according to the target focal length at the current position for image fusion can be executed as follows: First, determine the initial focal length subsequence at the current position according to the determined target focal length. The determination method can be to select near the target focal length according to the minimum focusing distance of the microscope lens. For example, select a continuous focal length sequence of the target focal length and the target focal length ± the minimum focusing distance to obtain an initial focal length subsequence of a preset length, and the preset length is at least set to 2. After determining the initial focal length subsequence, obtain the initial image subsequence collected by the microscope lens at the current position. The initial image subsequence is adapted to the initial focal length subsequence at the current position. Among them, the length of the obtained initial image subsequence is the preset length, which is the same as the number of the starting target image subsequences obtained in the above starting position scenario. In the embodiment of the present application, the length is at least 2. Then, based on a preset image clarity evaluation algorithm, calculate the clarity of each image in the initial image subsequence, and compare the clarity of each image with a preset clarity threshold for judgment. If the clarity value of each image is within the second preset threshold range, it is determined that the initial image subsequence is the target image subsequence at the current position; perform image fusion processing on the images in the target image subsequence to obtain the target clear image at the current position.

[0061] In the solution of the present application, according to the clear image subsequence obtained at the previous position and the height difference between the current position and the previous position, obtain the clear image of the back side of the wafer at the current position. In the preset length of the target image subsequence with higher clarity obtained at the previous position, there may be a height difference at the current position, but the height difference between the two positions is generally very small. Therefore, based on the sequence images obtained at the previous position, fine-tune the focal length to obtain the target image subsequence at the current position, which can save the time for obtaining images and the time for calculating the image clarity, thereby improving the efficiency of the image fusion process and obtaining a wafer back side image with higher clarity.

[0062] In one embodiment, considering that some images in the initial image subsequence collected by the microscope lens at the current position may be blurred, have low clarity, and do not meet the second preset threshold range, therefore, before performing image fusion processing on the images in the target image subsequence to obtain the target clear image at the current position, the following steps can also be executed:

[0063] If the clarity values of each image are not all within the second preset threshold range, that is, there is at least one image whose clarity is not within the second preset threshold range, then according to the image clarity calculation result and the minimum focusing distance of the microscope lens, the initial focal length subsequence at the current position is adjusted to generate an adjusted focal length subsequence at the current position. It can also be understood that, based on the image clarity calculation result, on the basis of the current initial focal length subsequence, the minimum focusing distance of the microscope lens is increased or decreased to generate a new adjusted focal length subsequence. Then, an adjusted image subsequence corresponding to the adjusted focal length subsequence at the current position is obtained, and it is re-determined whether the clarity values of the images in the adjusted image subsequence are all within the second preset threshold range. If the clarity values of the images in the adjusted image subsequence are all within the second preset threshold range, then the adjusted image subsequence is determined as the target image subsequence at the current position.

[0064] In the solution of this application, considering that there may be some images in the initial image subsequence collected by the microscope lens at the current position that are blurred, have low clarity, and do not meet the second preset threshold range, therefore, on the basis of the initially obtained initial focal length subsequence, fine-tuning is performed to obtain an adjusted focal length subsequence, and the corresponding target image subsequence is obtained according to the adjusted focal length subsequence, so as to ensure the clarity of the acquired images, and thus improve the quality of the fused images.

[0065] Figure 1 It is a schematic flowchart of an image processing method in an embodiment. It should be understood that although Figure 1 the steps in the flowchart are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows; unless there is a clear description in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders; and Figure 1 at least a part of the steps in

[0066] Based on the above method, an embodiment of this application also discloses an image processing device.

[0067] Referring to Figure 2 , the device includes the following modules:

[0068] A starting sequence determination module 201, which is used for any starting position on the back of the wafer to be collected, and determines the starting focal length sequence of the microscope lens according to the obtained minimum focusing distance of the microscope lens and the allowable acceptance height difference of the wafer;

[0069] A starting image acquisition module 202, configured to acquire a starting image sequence collected by a microscope lens at a starting position, where the starting image sequence is adapted to a starting focal length sequence;

[0070] A target image determination module 203, configured to calculate the clarity of each image in the starting image sequence based on a preset image clarity evaluation algorithm, determine an image subsequence according to a preset sequence length, and determine the image subsequence with the highest clarity score as the starting target image subsequence;

[0071] A clear image generation module 204, configured to perform image fusion processing on the images in the starting target image subsequence to obtain a target clear image at the starting position.

[0072] In one embodiment, the clear image generation module 204 is further configured to be applied to any non-starting position on the back of the wafer to be collected, determine the target focal length at the current position according to the target image subsequence at the previous position; acquire a first image collected by the microscope lens at the current position, where the first image is adapted to the target focal length; calculate the clarity of the first image based on a preset image clarity evaluation algorithm, determine whether the clarity of the first image is within a first preset threshold range, and determine the target clear image at the current position according to the determination result.

[0073] In one embodiment, the clear image generation module 204 is further configured to, if the clarity value of the first image is within the first preset threshold range, determine that the first image is the target clear image at the current position; if the clarity value of the first image is not within the first preset threshold range, determine an initial focal length subsequence at the current position according to the target focal length at the current position, and according to the current position's

[0074] In one embodiment, the clear image generation module 204 is further configured to acquire an initial image subsequence collected by the microscope lens at the current position, where the initial image subsequence is adapted to the initial focal length subsequence at the current position; calculate the clarity of each image in the initial image subsequence based on a preset image clarity evaluation algorithm, and if the clarity value of each image is within a second preset threshold range, determine that the initial image subsequence is the target image subsequence at the current position; perform image fusion processing on the images in the target image subsequence to obtain the target clear image at the current position.

[0075] In one embodiment, the clear image generation module 204 is further configured to, if the sharpness values of each image are not all within the second preset threshold range, adjust the initial focal length subsequence at the current position according to the calculated result of the image sharpness and the minimum focusing distance of the microscope lens, and generate an adjusted focal length subsequence at the current position; obtain an adjusted image subsequence corresponding to the adjusted focal length subsequence at the current position, and re-determine whether the sharpness values of the images in the adjusted image subsequence are all within the second preset threshold range. If the sharpness values of the images in the adjusted image subsequence are all within the second preset threshold range, it is determined that the adjusted image subsequence is the target image subsequence at the current position.

[0076] In one embodiment, the clear image generation module 204 is further configured to calculate the sharpness of each image in the target image subsequence at the previous position according to a preset image sharpness evaluation algorithm, and determine the focal length corresponding to the image with the highest sharpness; calculate the height difference between the current position and the previous position, and determine the target focal length at the current position according to the height difference and the focal length corresponding to the image with the highest sharpness at the previous position.

[0077] In one embodiment, the preset sequence length in the target image determination module 203 is at least 2.

[0078] The image processing apparatus provided in the embodiments of the present application can be applied to the image processing method provided in the above embodiments. For related details, refer to the above method embodiments. The implementation principles and technical effects are similar and will not be elaborated here.

[0079] It should be noted that when the image processing apparatus provided in the embodiments of the present application performs image processing, only the above-mentioned division of each functional module / functional unit is used for illustration. In actual applications, the above functions can be allocated to different functional modules / functional units according to needs, that is, the internal structure of the image processing apparatus is divided into different functional modules / functional units to complete all or part of the functions described above. In addition, the implementation manners of the image processing method provided in the above method embodiments and the implementation manners of the image processing apparatus provided in this embodiment belong to the same concept. For the specific implementation process of the image processing apparatus provided in this embodiment, refer to the above method embodiments and will not be elaborated here.

[0080] The embodiments of the present application also disclose a computer device.

[0081] Specifically, such as Figure 3As shown, the computer device can be a desktop computer, a laptop computer, a handheld computer, a cloud server, or other computer devices. The computer device may include, but is not limited to, a processor and a memory. Among them, the processor and the memory can be connected by a bus or other means. Among them, the processor can be a Central Processing Unit (CPU). The processor can also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, Graphics Processing Units (GPUs), Embedded Neural Network Processors (NPUs), or other dedicated deep learning coprocessors, discrete gate or transistor logic devices, discrete hardware components, etc. chips, or a combination of the above various types of chips.

[0082] As a non-transitory computer-readable storage medium, the memory can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the methods in the above embodiments of the present application. By running the non-transitory software programs, instructions, and modules stored in the memory, the processor can execute various functional applications and data processing of the processor, that is, implement the methods in the above method embodiments. The memory may include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created by the processor, etc. In addition, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory may optionally include a memory remotely provided relative to the processor, and these remote memories can be connected to the processor through a network. Examples of the above networks include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0083] The embodiments of the present application also disclose a computer-readable storage medium.

[0084] Specifically, a computer-readable storage medium is used to store a computer program. When the computer program is executed by a processor, the methods in the above method embodiments are implemented. Those skilled in the art can understand that to implement all or part of the processes in the above method embodiments of the present application, it can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above method embodiments. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above types of memories.

[0085] This specific embodiment is only an interpretation of the present invention, and it is not a limitation of the present invention. After reading this specification, those skilled in the art can make modifications to this embodiment without creative contributions as needed, but as long as it is within the scope of the claims of the present invention, it is protected by the patent law.

Claims

1. An image processing method, characterized in that: The method is applied to an automatic optical inspection system for wafer backside defects. The automatic optical inspection system for wafer backside defects includes a microscopic lens for collecting images of the wafer backside, and the focal length of the microscopic lens is adjustable. The method includes: Applied to any starting position of the wafer backside to be collected, determine the starting focal length sequence of the microscopic lens according to the obtained minimum focusing distance of the microscopic lens and the allowable height difference of the wafer; Obtain the starting image sequence collected by the microscopic lens at the starting position, where the starting image sequence corresponds to the starting focal length sequence; Based on a preset image sharpness evaluation algorithm, calculate the sharpness of each image in the starting image sequence, determine an image subsequence according to a preset sequence length, and determine the starting target image subsequence with the highest sharpness score; Perform image fusion processing on the images in the starting target image subsequence to obtain the target clear image at the starting position; After performing image fusion processing on the images in the starting target image subsequence to obtain the target clear image at the starting position, it further includes: Applied to any non-starting position of the wafer backside to be collected, determine the target focal length at the current position according to the target image subsequence at the previous position; Obtain the first image collected by the microscopic lens at the current position, where the first image corresponds to the target focal length; Based on a preset image sharpness evaluation algorithm, calculate the sharpness of the first image, determine whether the sharpness of the first image is within a first preset threshold range, and determine the target clear image at the current position according to the judgment result; The determining the target focal length at the current position according to the target image subsequence at the previous position includes: Calculate the sharpness of each image in the target image subsequence at the previous position according to a preset image sharpness evaluation algorithm, and determine the focal length corresponding to the image with the highest sharpness; Calculate the height difference between the current position and the previous position, and determine the target focal length at the current position according to the height difference and the focal length corresponding to the image with the highest sharpness at the previous position.

2. The method according to claim 1, wherein: The determining the target clear image at the current position according to the judgment result includes: If the sharpness value of the first image is within the first preset threshold range, then determine that the first image is the target clear image at the current position; If the sharpness value of the first image is not within the first preset threshold range, then determine the initial focal length subsequence at the current position according to the target focal length at the current position, and determine the target clear image at the current position according to the initial focal length subsequence at the current position.

3. The method according to claim 2, characterized in that: The determining the target clear image at the current position according to the initial focal length subsequence at the current position includes: Obtain the initial image subsequence collected by the microscopic lens at the current position, where the initial image subsequence corresponds to the initial focal length subsequence at the current position; Calculate the clarity of each image in the initial image subsequence based on a preset image clarity evaluation algorithm. If the clarity value of each image is within the second preset threshold range, determine that the initial image subsequence is the target image subsequence at the current position; Perform image fusion processing on the images in the target image subsequence to obtain the target clear image at the current position.

4. The method according to claim 3, wherein: Before performing the image fusion processing on the images in the target image subsequence to obtain the target clear image at the current position, it further includes: If the clarity value of each image is not within the second preset threshold range, adjust the initial focal length subsequence at the current position according to the image clarity calculation result and the minimum focusing distance of the microscope lens to generate the adjusted focal length subsequence at the current position; Obtain the adjusted image subsequence corresponding to the adjusted focal length subsequence at the current position, and re-determine whether the clarity values of the images in the adjusted image subsequence are all within the second preset threshold range. If the clarity values of the images in the adjusted image subsequence are all within the second preset threshold range, determine that the adjusted image subsequence is the target image subsequence at the current position.

5. The method according to claim 1, characterized in that: The preset sequence length is at least 2.

6. An image processing apparatus, characterized in that: The device is applied to an automatic optical inspection system for wafer backside defects. The automatic optical inspection system for wafer backside defects includes a microscope lens for collecting images of the wafer backside, and the focal length of the microscope lens is adjustable. The device includes: A start sequence determination module (201) for being applied to any start position of the wafer backside to be collected, and determining the start focal length sequence of the microscope lens according to the obtained minimum focusing distance of the microscope lens and the allowable height difference of the wafer; A start image acquisition module (202) for acquiring the start image sequence collected by the microscope lens at the start position, where the start image sequence is adapted to the start focal length sequence; A target image determination module (203) for calculating the clarity of each image in the start image sequence based on a preset image clarity evaluation algorithm, determining the image subsequence according to the preset sequence length, and determining the image subsequence with the highest clarity score as the start target image subsequence; A clear image generation module (204) for performing image fusion processing on the images in the start target image subsequence to obtain the target clear image at the start position; The clear image generation module (204) is further for being applied to any non-start position of the wafer backside to be collected. According to the target image subsequence at the previous position, determine the target focal length at the current position; acquire the first image collected by the microscope lens at the current position, where the first image is adapted to the target focal length; calculate the clarity of the first image based on a preset image clarity evaluation algorithm, determine whether the clarity of the first image is within the first preset threshold range, and determine the target clear image at the current position according to the determination result; The clear image generation module (204) is further configured to calculate the clarity of each image in the target image subsequence at the previous position according to a preset image clarity evaluation algorithm, and determine the focal length corresponding to the image with the highest clarity; calculate the height difference between the current position and the previous position, and determine the target focal length at the current position according to the height difference and the focal length corresponding to the image with the highest clarity at the previous position.

7. A computer device, characterized in that, It includes a memory and a processor, and a computer program capable of being loaded and executed by the processor, such as any one of the methods in claims 1 to 5, is stored on the memory.

8. A computer-readable storage medium, characterized in that, A computer program capable of being loaded and executed by the processor, such as any one of the methods in claims 1 to 5, is stored.

Images