Die bonder positioning method and system based on image recognition

Abstract

The invention relates to the technical field of image processing, in particular to a die bonder positioning method and system based on image recognition, and the method comprises the steps: collecting to-be-positioned chip images and standard chip images at different angles; segmenting the to-be-positioned chip images and the standard chip image to extract target areas in the to-be-positioned chip images and the standard chip image; identifying feature points of target areas of the to-be-positioned chip images and the standard chip image, and obtaining the number of target feature points of the feature points of the target areas of the to-be-positioned chip images outside the target area of the standard chip image; and calculating the matching degree between each to-be-positioned chip image and the standard chip image, and responding to the to-be-positioned chip image with the maximum matching degree as the optimal to-be-positioned chip image. According to the invention, the problem that the positioning precision is not high when the die bonder is used for positioning a chip is solved.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and more specifically, to a method and system for positioning a die bonder based on image recognition. Background Art

[0002] With the rapid development of the semiconductor industry, the requirements for production efficiency and precision in the die bonding process of chip packaging technology are constantly increasing. As an important equipment dedicated to chip packaging, the die bonding machine undertakes the key task of accurately bonding the chip to the substrate. The accuracy of the die bonding process directly affects the performance and reliability of the chip. Therefore, how to achieve high-precision chip positioning has become one of the core factors to improve the quality of the die bonding process.

[0003] With the rapid development of image processing and computer vision technology, positioning methods based on image recognition have begun to be widely used in the field of precision manufacturing. In the die bonding process, image recognition can not only achieve non-contact positioning, but also adjust parameters through software, reduce dependence on mechanical systems, and improve the flexibility and reliability of equipment. However, when dealing with chip images at different angles and under different lighting conditions, existing image recognition technology often cannot accurately extract feature areas, resulting in positioning errors. In addition, the image segmentation method based on fixed thresholds is easily disturbed when dealing with complex backgrounds, affecting the accurate recognition of the target area.

[0004] In traditional die-bonding positioning methods, mechanical guidance or optical detection systems are usually relied on. Although these methods can achieve certain positioning effects in some application scenarios, their positioning accuracy and efficiency are often limited due to the small size and complex shape of chips and substrates. Specifically, the mechanical guidance system has high requirements for equipment adjustment, and in the actual production process, equipment wear and environmental changes may cause positioning errors. In addition, under different lighting and angle conditions, the optical detection system is also easily affected by lighting conditions and environmental interference. These factors together lead to a decrease in the positioning accuracy of the die-bonding, making it difficult to further improve the positioning accuracy during the die-bonding process. Summary of the invention

[0005] In order to solve the problem in the background art that it is difficult to further improve the positioning accuracy during the die bonding process, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a method for positioning a die bonder based on image recognition, comprising: collecting images of each chip to be positioned and a standard chip image at different angles; segmenting each chip image to be positioned and the standard chip image respectively to extract a target area in the chip image to be positioned and the standard chip image; wherein the process of segmenting each chip image to be positioned and the standard chip image respectively comprises: setting an initial local threshold of an adaptive threshold binarization algorithm, segmenting each chip image to be positioned and the standard chip image respectively, and updating the initial local threshold, performing multiple segmentations, obtaining a change in the area of ​​a connected domain before and after each segmentation, and taking the maximum change as the segmentation value. The connected domain corresponding to the value is taken as the target area; the feature points of the target areas of each chip image to be located and the standard chip image are identified, and the transformed positions of the feature points of the target areas of each chip image to be located in the standard chip image are obtained, and the number of target feature points of the target areas of each chip image to be located that fall outside the target area of ​​the standard chip image is obtained based on the transformed positions; the matching degree of each chip image to be located and the standard chip image is calculated, and the matching degree is inversely correlated with the number of chip images to be located and the number of feature points of the target areas of each chip image to be located, and is positively correlated with the number of target feature points; the chip image to be located with the largest matching degree is the optimal chip image to be located.

[0007] The above technical solution collects the image of the chip to be located and the standard chip image from multiple angles, uses an adaptive threshold binarization algorithm to perform multiple segmentations and connected domain area change analysis, accurately extracts the target area, and combines the matching of feature points and the calculation of the transformed position to ensure high-precision matching of the chip image. It is also suitable for different types of chips and has significant technical advantages in the chip manufacturing and packaging process.

[0008] Furthermore, the matching degree p between the i-th chip image to be positioned and the standard chip image is i for:

[0009]

[0010] Where n represents the number of images of the i-th chip to be positioned at different angles, c i,j represents the number of feature points in the target area of ​​the image of the jth angle of the i-th chip to be positioned, c i ′ ,j It indicates the number of target feature points whose feature points in the target area of ​​the i-th image of the chip to be positioned at the j-th angle fall outside the target area of ​​the standard chip image.

[0011] The above technical solution realizes the accurate matching evaluation of the chip to be positioned at different angles and the standard chip image, comprehensively considers multi-angle information, and improves the accuracy and robustness of the matching process. The matching degree quantifies the similarity between chip images by calculating the deviation ratio of feature points, so that the optimal chip image to be positioned can be efficiently screened, which significantly improves the automation and recognition accuracy of chip image matching.

[0012] Furthermore, a CMOS camera is used to collect images of the chips to be positioned and the standard chip images at different angles.

[0013] Furthermore, the method further includes performing image grayscale processing and median filtering processing on the images of the chip to be positioned at different angles and the images of the standard chip at a standard angle.

[0014] Furthermore, the updating of the initial local threshold is specifically:

[0015] Set the update range of the initial local threshold to [d 0 ,d 1 ], the update rule is: the last local threshold plus d is used as the local threshold after the current update, d 0 represents the initial local threshold, d 1 Represents the termination local threshold.

[0016] The above technical solution realizes precise control of adaptive threshold adjustment, effectively improves the flexibility and accuracy of the segmentation process, ensures that the connected domain corresponding to the maximum change is captured in multiple segmentations, thereby more accurately extracting the target area and improving the efficiency and robustness of image segmentation.

[0017] Furthermore, the changing position is specifically:

[0018] A feature point detection algorithm is used to extract feature points of the target area of ​​each chip image to be located and the standard chip image, and the feature points of the target area of ​​each chip image to be located are matched with the feature points of the target area of ​​the standard chip image through a nearest neighbor search and matching algorithm; based on the matching result, a geometric transformation matrix between the feature points of the target area of ​​each chip image to be located and the feature points of the target area of ​​the standard chip image is calculated; using the geometric transformation matrix, the feature points of each chip image to be located are converted to the target area of ​​the standard chip image, and then the transformed position of the feature points of each chip image target area to be located in the target area of ​​the standard chip image is obtained.

[0019] The above technical solution uses the feature point detection algorithm and the nearest neighbor search and matching algorithm to accurately extract and match the feature points of the chip to be located and the standard chip, calculate their geometric transformation matrix, and realize the spatial transformation of the feature points. By converting the feature points of the image of the chip to be located to the target area of ​​the standard chip image, the transformation position of the feature points can be accurately obtained, thereby effectively improving the accuracy and robustness of image matching, and helping to achieve efficient and accurate chip positioning and matching.

[0020] Furthermore, the feature point detection algorithm includes: SIFT, SURF or ORB algorithm.

[0021] In a second aspect, the present invention provides a die bonding machine positioning system based on image recognition, comprising a memory and a processor, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, any one of the above-described methods for positioning a die bonding machine based on image recognition is implemented.

[0022] The beneficial effects of the present invention are:

[0023] The present invention segments the chip image to be located and the standard chip image, dynamically adjusts the threshold value in combination with the adaptive threshold binarization algorithm, extracts the target area, and calculates the image matching degree based on feature point matching and geometric transformation. The matching degree is measured according to the number and distribution of feature points, ensuring the accurate comparison between the chip to be located and the standard chip. This process can effectively reduce errors, improve the accuracy of image segmentation and feature matching, and finally select the optimal chip image to be located through the matching degree, so as to achieve efficient and accurate positioning of the chip image. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0025] Figure 1 is a flow chart schematically showing a method for positioning a die bonding machine based on image recognition according to an embodiment of the present invention;

[0026] Figure 2 FIG. 4 is a block diagram schematically showing a structure of a die bonding machine positioning system based on image recognition according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0028] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0029] An embodiment of a method for positioning a die bonding machine based on image recognition.

[0030] like Figure 1 As shown, a flow chart of a method for positioning a die bonding machine based on image recognition according to an embodiment of the present invention includes the following steps:

[0031] S1: Collect images of the chips to be positioned and the standard chip images at different angles.

[0032] In one embodiment, a high-resolution CMOS camera may be used to collect images of the chips to be positioned and the standard chip at different angles. Of course, a CCD camera or other cameras may also be used to collect images of the chips to be positioned and the standard chip.

[0033] Specifically, the chip to be positioned is fixed on the positioning device, the angle change is controlled by a mechanical arm or a rotating table, and the shooting angle of the camera is adjusted successively to obtain images of the chip to be positioned observed from multiple different angles, and the standard chip image is taken at the same time.

[0034] In this embodiment, after collecting the chip image, the images of each chip to be located and the standard chip image are first preprocessed. The image preprocessing process includes image grayscale processing and median filtering processing. For grayscale processing: convert the color image into a grayscale image, remove redundant color information, and simplify the subsequent image processing process. The grayscale process can use a method based on RGB color channel weighting to calculate the grayscale value of each pixel. The grayscale image can more intuitively present the boundaries and structural features of the chip, laying the foundation for the subsequent extraction and matching of feature points. For median filtering processing: eliminate salt and pepper noise while maintaining the edge details of the image. The steps of median filtering processing are: for each pixel point of the image, set a neighborhood window around it, calculate the median of the pixel value in the window, and replace the original pixel value with the median. This process can significantly reduce image noise, improve image quality, and ensure that the chip features are clearly presented in the image.

[0035] S2: Segmenting each chip image to be positioned and the standard chip image respectively to extract the target area in each chip image to be positioned and the standard chip image.

[0036] In one embodiment, the process of segmenting each chip image to be positioned and the standard chip image is as follows: setting the initial local threshold of the adaptive threshold binarization algorithm, segmenting each chip image to be positioned and the standard chip image, and updating the initial local threshold, performing multiple segmentations, obtaining the change in the area of ​​the connected domain before and after each segmentation, and taking the connected domain corresponding to the maximum value of the change as the target area;

[0037] The specific method for updating the initial local threshold is as follows: setting the update range of the initial local threshold to [d 0 ,d 1 ], the update rule is: the last local threshold plus d is used as the local threshold after the current update, d 0 represents the initial local threshold, d 1 Represents the termination local threshold.

[0038] For example, assume that a high-resolution CMOS camera is used to capture five images of the chip to be positioned at different angles and one image of a standard chip, and an adaptive threshold binarization algorithm is used for initial segmentation. According to the image brightness distribution, the initial local threshold range is set to [100, 150], d is 5, and it is assumed that the image is segmented with an initial local threshold of 100 to obtain a connected domain area M, and the initial local threshold is updated to 105 to segment the image to obtain a connected domain area N, then the updated connected domain area change is MN, and the initial local threshold is gradually updated to obtain the change in the connected domain area each time, and the connected domain corresponding to the maximum change is taken as the target area.

[0039] S3: Identify the feature points of the target area of ​​each chip image to be positioned and the standard chip image, and obtain the transformed position of the feature points of the target area of ​​each chip image to be positioned in the standard chip image.

[0040] In one embodiment, obtaining the transformed position of the feature points of each target area of ​​the chip image to be located in the standard chip image is specifically as follows: extracting the feature points of each target area of ​​the chip image to be located and the standard chip image using a feature point detection algorithm, and matching the feature points of the target area of ​​the chip image to be located with the feature points of the target area of ​​the standard chip image using a nearest neighbor search matching algorithm; calculating the geometric transformation matrix between the feature points of the target area of ​​the chip image to be located and the feature points of the target area of ​​the standard chip image based on the matching result; converting the feature points of each target area of ​​the chip image to be located into the target area of ​​the standard chip image using the geometric transformation matrix, and then obtaining the transformed position of the feature points of each target area of ​​the chip image to be located in the target area of ​​the standard chip image. In this embodiment, the feature point detection algorithm used is the SIFT algorithm.

[0041] In another embodiment, the feature point detection algorithm includes: SIFT, SURF or ORB algorithm. The SIFT algorithm has strong scale and rotation invariance and is suitable for target matching in complex backgrounds; the SURF algorithm is faster than SIFT and is suitable for application scenarios with high real-time requirements; the ORB algorithm is an efficient feature detection and matching algorithm that can reduce computational complexity while ensuring accuracy and is more suitable for use in resource-constrained devices.

[0042] S4: Calculating the matching degree between each chip image to be positioned and the standard chip image based on the changed position, and responding that the chip image to be positioned with the greatest matching degree is the optimal chip image to be positioned.

[0043] In one embodiment, the matching degree p of the i-th chip image to be positioned and the standard chip image is i for:

[0044]

[0045] Where n represents the number of images of the i-th chip to be positioned at different angles, c i,j Indicates the number of feature points in the target area of ​​the image of the jth angle of the i-th chip to be positioned, c′ i,j It indicates the number of target feature points whose feature points in the target area of ​​the i-th image of the chip to be positioned at the j-th angle fall outside the target area of ​​the standard chip image.

[0046] For example, assume that the system collects 3 images of the chip to be positioned at different angles and matches them with the standard chip image. For the i-th image of the chip to be positioned, the number of images at different angles is 3. We assume the following data: For the image at the first angle, the number of feature points in the target area c 1,1 is 50, and the number of feature points outside the target area of ​​the standard chip image is c 1 ′ ,1 is 5; for the image at the second angle, the number of feature points in the target area c 1,2 is 60, and the number of feature points outside the target area of ​​the standard chip image is c 1 ′ ,2 is 10; for the image at the third angle, the number of feature points in the target area c 1,3 is 55, and the number of feature points outside the target area of ​​the standard chip image is c 1 ′ ,3 is 8; substituting it into the matching degree formula, the matching degree between the i-th chip image to be located and the standard chip image can be obtained.

[0047] The matching degrees of all chip images to be located are obtained, and the chip image to be located with the largest matching degree is selected as the optimal chip image to be located.

[0048] The solution of the present invention collects images of the chip to be located and the standard chip images at different angles, and uses an adaptive threshold binarization algorithm for segmentation to effectively extract the feature points of the target area, thereby improving the accuracy of image processing. By calculating the transformed position of the feature points of the target area and their matching degree in the standard chip image, the matching effect of each chip image to be located is evaluated, and then the best chip image to be located is identified, thereby optimizing the efficiency of chip positioning and identification. This process not only enhances the reliability of feature point detection, but also improves the accuracy of the matching algorithm, providing a more effective solution for practical applications.

[0049] An embodiment of a die bonding machine positioning system based on image recognition:

[0050] like Figure 2 As shown, a structural block diagram of a die bonding machine positioning system based on image recognition according to an embodiment of the present invention includes a processor and a memory.

[0051] The present invention also provides a die bonding machine positioning system based on image recognition. Figure 2 As shown, the system includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a die bonding machine positioning method based on image recognition according to the present invention is implemented.

[0052] The image recognition-based die bonding machine positioning system also includes other components well known to those skilled in the art, such as a communication interface, and the configuration and functions of the components are known in the art, so they will not be described in detail here.

[0053] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that may be used to store the required information and may be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented using computer-readable / executable instructions that may be stored or otherwise maintained by such a computer-readable medium.

[0054] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0055] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A method for positioning a die bonder based on image recognition, characterized in that: include: Collect images of each chip to be positioned and images of the standard chip at different angles; Segmenting each chip image to be positioned and the standard chip image respectively to extract the target area in each chip image to be positioned and the standard chip image; The process of segmenting each chip image to be positioned and the standard chip image respectively is as follows: setting the initial local threshold of the adaptive threshold binarization algorithm, segmenting each chip image to be positioned and the standard chip image respectively, and updating the initial local threshold, performing multiple segmentations, obtaining the change in the area of ​​the connected domain before and after each segmentation, and taking the connected domain corresponding to the maximum value of the change as the target area; identifying the feature points of the target area of ​​each chip image to be positioned and the standard chip image, and obtaining the transformed position of the feature points of the target area of ​​each chip image to be positioned in the standard chip image, and obtaining the number of target feature points of the target area of ​​each chip image to be positioned that fall outside the target area of ​​the standard chip image based on the transformed position; calculating the matching degree of each chip image to be positioned and the standard chip image, wherein the matching degree is inversely correlated with the number of each chip image to be positioned and the number of feature points of the target area of ​​each chip image to be positioned, and is positively correlated with the number of target feature points; The chip image to be positioned with the largest matching degree is regarded as the optimal chip image to be positioned.

2. The method for positioning a die bonder based on image recognition according to claim 1, characterized in that: The matching degree p of the i-th chip image to be positioned and the standard chip image i for: Where n represents the number of images of the i-th chip to be positioned at different angles, c i,j represents the number of feature points in the target area of ​​the image of the jth angle of the i-th chip to be positioned, c i ′ ,j It indicates the number of target feature points whose feature points in the target area of ​​the i-th image of the chip to be positioned at the j-th angle fall outside the target area of ​​the standard chip image.

3. The method for positioning a die bonder based on image recognition according to claim 1, characterized in that: The CMOS camera is used to collect images of the chips to be positioned and the standard chip images at different angles.

4. The method for positioning a die bonding machine based on image recognition according to claim 1, characterized in that: The method also includes performing image grayscale processing and median filtering processing on the images of the chip to be positioned at different angles and the images of the standard chip at a standard angle.

5. The method for positioning a die bonder based on image recognition according to claim 1, characterized in that: The updating initial local threshold is specifically: The update range of the initial local threshold is set to [d0, d1], and the update rule is: the last local threshold plus d is used as the local threshold after the current update, d0 represents the initial local threshold, and d1 represents the termination local threshold.

6. The method for positioning a die bonder based on image recognition according to claim 1, characterized in that: The change position is specifically: Extracting feature points of the target area of ​​each chip image to be located and the standard chip image by using a feature point detection algorithm, and matching the feature points of the target area of ​​each chip image to be located with the feature points of the target area of ​​the standard chip image by using a nearest neighbor search and matching algorithm; Based on the matching results, a geometric transformation matrix between the feature points of each target area of ​​the chip image to be located and the feature points of the target area of ​​the standard chip image is calculated; The geometric transformation matrix is ​​used to transform the feature points of each chip image to be positioned into the target area of ​​the standard chip image, thereby obtaining the transformed position of the feature points of each chip image target area to be positioned in the target area of ​​the standard chip image.

7. The method for positioning a die bonder based on image recognition according to claim 6, characterized in that: The feature point detection algorithm includes: SIFT, SURF or ORB algorithm.

8. A die bonding machine positioning system based on image recognition, characterized in that: The invention comprises a memory and a processor, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a method for positioning a die bonding machine based on image recognition as claimed in any one of claims 1 to 7 is implemented.

Images