Appearance detection equipment for chip production
By using appearance inspection equipment in chip production, infrared images are acquired and preprocessed to extract the detection area and calculate the infrared temperature value, thus solving the problem of insufficient accuracy in infrared imaging detection and achieving efficient and accurate chip appearance inspection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SUZHOU SHANGZHUN ELECTRONIC TECH CO LTD
- Filing Date
- 2026-03-23
- Publication Date
- 2026-05-12
AI Technical Summary
Existing technologies struggle to make detailed judgments during infrared imaging detection in chip manufacturing, resulting in insufficient detection accuracy and a high risk of errors.
Using appearance inspection equipment for chip manufacturing, infrared images and chip information of the target chip are acquired through the acquisition component. The detection component is used for preprocessing, the detection area is extracted, the infrared temperature value of unit pixel is calculated, an integrity index is created, and chips with unqualified appearance are screened out.
It improves the accuracy and efficiency of chip appearance inspection, ensures the accuracy and reliability of inspection results, and avoids the influence of mechanical errors and environmental factors.
Smart Images
Figure CN122016932A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of chip inspection technology, and more specifically, to an appearance inspection device for chip manufacturing. Background Technology
[0002] In electronics, a chip is a way to miniaturize circuits (mainly semiconductor devices, but also passive components, etc.) and is often manufactured on the surface of a semiconductor wafer.
[0003] Integrated circuits fabricated on the surface of semiconductor chips are also known as thin-film integrated circuits. Another type, thick-film integrated circuits (hybrid integrated circuits), are miniaturized circuits composed of independent semiconductor devices and passive components integrated onto a substrate or circuit board.
[0004] Chinese Patent CN119643585B discloses a chip defect detection system and method based on visual recognition, including: S10: partitioning the surface of an epitaxial wafer, with the internal region of each closed edge line in the grain boundary image corresponding to a partition; S20: predicting the defect index of each partition; S30: predicting the resistivity uniformity of each partition; S40: screening epitaxial wafers with defects. This invention evaluates the production quality of epitaxial wafers through multiple physical characteristics, further improving the system's defect detection accuracy. However, in existing technologies, it is difficult to make detailed judgments when using infrared imaging, leading to insufficient judgment accuracy and making the final detection results prone to errors. Summary of the Invention
[0005] The purpose of this invention is to provide an appearance inspection device for chip manufacturing, so as to solve the problem of insufficient inspection accuracy mentioned in the background art.
[0006] To address the aforementioned technical problems, one objective of this invention is to provide an appearance inspection device for chip manufacturing, comprising an acquisition component and a detection component. The acquisition component acquires infrared images and chip information of multiple target chips. The detection component is communicatively connected to the acquisition component. All infrared images acquired by the acquisition component are input into the detection component, which then performs the following detection steps: All infrared images are preprocessed to obtain multiple preprocessed images; The detection area of each target chip is extracted based on the preprocessed image; The infrared temperature value corresponding to a unit pixel of the target chip is calculated by combining the detection area, and the integrity index of the target chip is created. Based on the integrity index of the target chip, analyze the appearance integrity of the chip in real time and screen out abnormal chips that fail to meet the appearance integrity standard.
[0007] Preferably, the acquisition component includes a chip acquisition module and an image acquisition module. The infrared image of the target chip is acquired through the temperature acquisition module, and the chip information of the target chip is acquired through the chip acquisition module.
[0008] Preferably, the detection component includes a preprocessing module and a detection module. The preprocessing module performs preprocessing on the infrared image and extracts the detection area. The detection module judges the appearance integrity of the target chip based on the detection area.
[0009] Preferably, all infrared images are preprocessed to obtain multiple preprocessed images, including: Randomly select one infrared image from all the infrared images; The infrared image is processed by Gaussian filtering to obtain a Gaussian filtered image; The Gaussian filtered image is subjected to histogram equalization to obtain a preprocessed image; Return to randomly select one infrared image from all infrared images, and continue until all infrared images have been selected, resulting in multiple preprocessed images.
[0010] Preferably, the detection region of each target chip is extracted based on the preprocessed image, including: Select a preprocessed image; The preprocessed image is then normalized to obtain a normalized image; The Sobel edge operator is used to detect the normalized image to obtain the edge pixels of the target chip; By sequentially connecting multiple edge pixels, the detection area of the target chip is obtained; Return to select a preprocessed image, and continue until all preprocessed images have been selected to obtain the detection area for each target chip.
[0011] Preferably, the infrared temperature value corresponding to a unit pixel of the target chip is calculated based on the detection area, and an integrity index of the target chip is created, including: Select a target chip and obtain the detection area of the target chip; Calculate the detection index of the detection area of the target chip; Establish an integrity index for the target chip based on detection indicators; Return to the selected target chip and obtain the detection area of the target chip, until all target chips have been selected, and obtain the integrity index of each target chip.
[0012] Preferably, the detection indicators of the detection area of the target chip are calculated, including: Count the number of pixels contained in the detection area of the target chip; Obtain the infrared temperature at each observation point of the target chip; The standard infrared temperature per pixel is calculated by combining the infrared temperature of the target chip with the number of pixels; the standard infrared temperature per pixel is the detection index of the target chip.
[0013] Preferably, the integrity index of the target chip is established based on the total detection index, including: Obtain the detection index corresponding to each infrared image of the target chip; Calculate the average value of all detection indicators; this average value is the integrity index of the target chip.
[0014] Preferably, after acquiring the infrared image of the target chip, it is necessary to perform image thermal compensation on the infrared image based on the surface temperature of the target chip.
[0015] Preferably, the operating temperature range of the acquisition component is set to [33 degrees Celsius - 36 degrees Celsius].
[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: Infrared images and chip information of different target chips are acquired by the acquisition component, and all acquired data are input into the detection component. The detection component performs preprocessing of the infrared images to obtain multiple preprocessed images. Then, the detection area of each target chip is extracted based on the preprocessed images, and the infrared temperature value corresponding to each pixel of the target chip is calculated based on the detection area. A target chip integrity index is created, and finally, the appearance integrity of the chip in real time is analyzed based on the target chip integrity index, and abnormal chips with unqualified appearance integrity are screened out. This application extracts the detection area of the target chip and creates an integrity index by combining the number of pixels in the detection area with the infrared temperature value, thereby coupling the appearance integrity of the target chip with infrared temperature, thus improving the detection efficiency of appearance integrity. Attached Figure Description
[0017] Figure 1 A connection diagram for appearance inspection equipment used in chip manufacturing; Figure 2 A flowchart illustrating the inspection components of an appearance inspection equipment used in chip manufacturing; Reference numerals: 100, Acquisition component; 101, Chip acquisition module; 102, Image acquisition module; 200. Detection component; 201. Preprocessing module; 202. Detection module. Detailed Implementation
[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0019] Example 1 like Figure 1 As shown, one of the objectives of this invention is to provide an appearance inspection device for chip manufacturing, including an acquisition component 100 and a detection component 200. The acquisition component 100 acquires infrared images and chip information of multiple target chips. The detection component 200 is communicatively connected to the acquisition component 100. All infrared images acquired by the acquisition component 100 are input into the detection component 200, and the detection component 200 performs the following detection steps: S100 preprocesses all infrared images to obtain multiple preprocessed images; S200 extracts the detection area of each target chip based on the preprocessed image; S300 calculates the infrared temperature value corresponding to a unit pixel of the target chip by combining the detection area, and creates an integrity index of the target chip. S400 analyzes the appearance integrity of the target chip in real time based on the chip integrity index and filters out abnormal chips that fail to meet the appearance integrity standard.
[0020] It should be noted that the acquisition component 100 acquires infrared images and chip information of different target chips, and inputs all acquired data into the detection component 200. The detection component 200 performs preprocessing of the infrared images to obtain multiple preprocessed images. Then, based on the preprocessed images, the detection area of each target chip is extracted, and the infrared temperature value corresponding to a unit pixel of the target chip is calculated in combination with the detection area. A target chip integrity index is then created. Finally, the real-time appearance integrity of the chip is analyzed based on the target chip integrity index, and abnormal chips with unqualified appearance integrity are screened out. This application extracts the detection area of the target chip and creates an integrity index by combining the number of pixels in the detection area with the infrared temperature value, thereby coupling the appearance integrity of the target chip with the infrared temperature, thus improving the detection efficiency of appearance integrity.
[0021] In one embodiment of this application, the acquisition component 100 includes a chip acquisition module 101 and an image acquisition module 102. The image acquisition module 102 acquires infrared images of the target chip, and the chip acquisition module 101 acquires chip information of the target chip.
[0022] It should be noted that different data information of the target chip is collected by different acquisition components to avoid information leakage due to information overlap. In order to increase the number of training samples and ensure data reliability, multiple infrared images of the target chip at different times need to be collected when collecting data for each target chip. This ensures the reliability of the data source when creating the integrity index of the target chip in the later stage.
[0023] In one embodiment of this application, the detection component 200 includes a preprocessing module 201 and a detection module 202. The preprocessing module 201 performs preprocessing on the infrared image and extracts the detection area. The detection module 202 judges the appearance integrity of the target chip based on the detection area.
[0024] It should be noted that chip appearance inspection is a key step in ensuring chip quality and performance, and timely detection of chip surface defects, dimensional deviations and other issues. Therefore, the preprocessing module 201 preprocesses all infrared images to improve image quality, so that the subsequent detection module 202 can make judgments based on high-quality training data when inspecting the appearance integrity of the target chip, thereby improving detection efficiency.
[0025] In one embodiment of this application, S100 includes: S110, randomly select one infrared image from all infrared images; S120, Gaussian filtering is applied to the infrared image to obtain a Gaussian filtered image; S130, Histogram equalization is performed on the Gaussian filtered image to obtain the preprocessed image; S140, return to randomly select one infrared image from all infrared images, until all infrared images have been selected, resulting in multiple preprocessed images.
[0026] It should be noted that this application uses infrared detection to determine the appearance integrity of the target chip. Based on the different absorption and emission characteristics of infrared radiation in different parts of the target chip, potential problems inside or on the surface of the chip are discovered by detecting the infrared image of the target chip. Some appearance defects may cause abnormal local temperature distribution of the chip, which will be displayed in the infrared image, thereby completing the detection of the appearance integrity of the target chip.
[0027] To improve the quality of infrared images, Gaussian filtering is used to remove high-frequency noise from the infrared images, thereby avoiding the influence of noise on the detection results. Histogram equalization, on the other hand, stretches the grayscale distribution to improve the global contrast of the infrared image.
[0028] In one embodiment of this application, S200 includes: S210, Select a preprocessed image; S220, normalize the preprocessed image to obtain a normalized image; S230, Perform Sobel edge operator detection on the normalized image to obtain the edge pixels of the target chip; S240, connects multiple edge pixels in sequence to obtain the detection area of the target chip; S250, return to select a preprocessed image, until all preprocessed images have been selected, and obtain the detection area of each target chip.
[0029] It should be noted that Sobel edge detection is an important processing method in the field of computer vision. It is mainly used to obtain the first-order gradient of digital images. Its common application and physical meaning is edge detection. Sobel edge detection calculates the weighted difference of the gray values of the four neighborhoods above, below, left, and right of each pixel in the image, and reaches the extreme value at the edge to detect the edge.
[0030] This application extracts the detection area and background area of the target chip from the preprocessed image of the target chip. The background area is the area other than the detection area in the preprocessed image. By extracting the detection area, on the one hand, the amount of data that needs to be processed in the detection process is reduced, thereby improving the detection efficiency. On the other hand, the detection area itself is one of the indicators for judging the appearance integrity of the target chip. If a target chip obtains multiple detection areas after edge detection, it proves that an abnormality may have occurred in the production process. Therefore, it can be determined that the appearance integrity of the target chip is unqualified.
[0031] In one embodiment of this application, S300 includes: S310, Select a target chip and obtain the detection area of the target chip; S320, calculate the detection index of the detection area of the target chip; S330 establishes integrity indicators for the target chip based on detection metrics; S340, return to select a target chip and obtain the detection area of the target chip, until all target chips have been selected, and obtain the integrity index of each target chip.
[0032] It should be noted that the detection index is actually a ratio describing the infrared temperature within the target chip to the number of pixels. Generally speaking, since it remains stable during the chip manufacturing process, chips of the same model will have the same detection index after processing. If the detection index of a certain target chip differs significantly from that of other target chips of the same model, it can be considered that the appearance integrity of the target chip is abnormal, thereby improving the detection accuracy.
[0033] The integrity index is obtained by averaging the test indicators of multiple target chips of the same model. The reason for using the average value is to avoid the influence of mechanical errors in the production process on the judgment of the final test results.
[0034] In one embodiment of this application, S320 includes: S321, count the number of pixels contained in the detection area of the target chip; S322, obtain the infrared temperature of each observation point of the target chip; Specifically, the observation points are random points set on the target chip, and the positions of the observation points can be determined before acquiring infrared images; S323, calculate the standard infrared temperature per unit pixel by combining the infrared temperature of the target chip and the number of pixels; the standard infrared temperature per unit pixel is the detection index of the target chip.
[0035] It should be noted that the standard infrared temperature per unit pixel is calculated using a formula; formula; in, It is the standard infrared temperature per pixel. is the infrared temperature of the i-th observation point, N is the total number of observation points, and P is the number of pixels contained in the detection area.
[0036] By converting the integrity of the target chip into a standard infrared temperature per pixel, the goal of judging the appearance integrity of the target chip based on changes in infrared temperature is achieved, thus realizing non-contact detection and avoiding any impact on chip production.
[0037] S330 in one embodiment of this application includes: S331, acquire the detection index corresponding to each infrared image of the target chip; S332, calculate the average value of all detection indicators, which is the integrity index of the target chip.
[0038] It should be noted that after obtaining the integrity index of each target chip, the real-time appearance integrity of the chip can be judged by setting an error threshold. The specific steps are as follows: S410 acquires real-time infrared images and real-time chip information from the chip. S420 calculates real-time detection metrics of the real-time chip based on real-time infrared images and real-time chip information. Specifically, the real-time detection index is actually the ratio of the infrared temperature in the detection area of the real-time chip to the number of pixels, and its calculation process is the same as that of the target chip's detection index. S430 performs cluster analysis on real-time chips based on real-time chips to obtain the target chips corresponding to the real-time chips; the target chips corresponding to the real-time chips are denoted as standard chips. Specifically, cluster analysis can be performed using the K-means clustering method; S440: Obtain the integrity index of the standard chip and calculate the difference between the integrity index of the standard chip and the real-time detection index of the real-time chip. S450 determines whether the difference between the integrity index of the standard chip and the real-time detection index of the real-time chip is greater than or equal to the difference threshold. S460, if the difference between the integrity index of the standard chip and the real-time detection index of the real-time chip is greater than or equal to the difference threshold, then the appearance integrity of the real-time chip is unqualified.
[0039] In one embodiment of this application, after acquiring the infrared image of the target chip, it is necessary to perform image thermal compensation on the infrared image based on the surface temperature of the target chip.
[0040] It should be noted that infrared imaging relies on the infrared radiation emitted by an object. The intensity of radiation depends not only on temperature but also on factors such as ambient temperature, distance, and viewing angle. Without thermal compensation, the temperature difference on the chip surface may be misinterpreted as a defect or mask the real defect. Therefore, image thermal compensation is used to eliminate or correct image errors caused by environmental factors, equipment drift, emissivity differences, and other non-target temperature changes, so that the infrared image of the target chip can more accurately reflect the true temperature distribution of the target chip under test, thereby improving the reliability of defect detection.
[0041] In one embodiment of this application, the operating temperature range of the acquisition component 100 is set to [33 degrees Celsius - 36 degrees Celsius].
[0042] It should be noted that by setting the operating temperature range, the efficiency of the acquisition component 100 is ensured, and abnormalities in the final acquired data are avoided due to excessively high or low operating temperatures causing malfunctions in the acquisition component.
[0043] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.
Claims
1. A visual inspection device for chip manufacturing, characterized in that, include: Data acquisition components; The acquisition component acquires infrared images and chip information of multiple target chips. Detection components; The detection component is communicatively connected to the acquisition component. All infrared images acquired by the acquisition component are input into the detection component, and the detection component performs the following detection steps: All infrared images are preprocessed to obtain multiple preprocessed images; The detection area of each target chip is extracted based on the preprocessed image; The infrared temperature value corresponding to a unit pixel of the target chip is calculated by combining the detection area, and the integrity index of the target chip is created. Based on the integrity index of the target chip, analyze the appearance integrity of the chip in real time and screen out abnormal chips that fail to meet the appearance integrity standard.
2. The appearance inspection equipment for chip manufacturing according to claim 1, characterized in that, The acquisition component includes a chip acquisition module and an image acquisition module. The image acquisition module acquires infrared images of the target chip, and the chip acquisition module acquires chip information of the target chip.
3. The appearance inspection equipment for chip manufacturing according to claim 2, characterized in that, The detection component includes a preprocessing module and a detection module. The preprocessing module performs preprocessing on the infrared image and extracts the detection area. The detection module judges the appearance integrity of the target chip based on the detection area.
4. The appearance inspection equipment for chip manufacturing according to claim 3, characterized in that, All infrared images were preprocessed to obtain multiple preprocessed images, including: Randomly select one infrared image from all the infrared images; The infrared image is processed by Gaussian filtering to obtain a Gaussian filtered image; The Gaussian filtered image is subjected to histogram equalization to obtain a preprocessed image; Return to randomly select one infrared image from all infrared images, and continue until all infrared images have been selected, resulting in multiple preprocessed images.
5. The appearance inspection equipment for chip manufacturing according to claim 4, characterized in that, The detection region for each target chip is extracted based on the preprocessed image, including: Select a preprocessed image; The preprocessed image is then normalized to obtain a normalized image; The Sobel edge operator is used to detect the normalized image to obtain the edge pixels of the target chip; By sequentially connecting multiple edge pixels, the detection area of the target chip is obtained; Return to select a preprocessed image, and continue until all preprocessed images have been selected to obtain the detection area for each target chip.
6. The appearance inspection equipment for chip manufacturing according to claim 5, characterized in that, The infrared temperature value corresponding to each pixel of the target chip is calculated based on the detection area, and an integrity index of the target chip is created, including: Select a target chip and obtain the detection area of the target chip; Calculate the detection index of the detection area of the target chip; Establish an integrity index for the target chip based on detection indicators; Return to the selected target chip and obtain the detection area of the target chip, until all target chips have been selected, and obtain the integrity index of each target chip.
7. The appearance inspection equipment for chip manufacturing according to claim 6, characterized in that, Calculate the detection metrics for the detection area of the target chip, including: Count the number of pixels contained in the detection area of the target chip; Obtain the infrared temperature at each observation point of the target chip; The standard infrared temperature per pixel is calculated by combining the infrared temperature of the target chip with the number of pixels; the standard infrared temperature per pixel is the detection index of the target chip.
8. The appearance inspection equipment for chip manufacturing according to claim 7, characterized in that, The integrity index of the target chip is established based on the total detection index, including: Obtain the detection index corresponding to each infrared image of the target chip; Calculate the average value of all detection indicators; this average value is the integrity index of the target chip.
9. The appearance inspection equipment for chip manufacturing according to claim 8, characterized in that, After acquiring the infrared image of the target chip, it is necessary to perform image thermal compensation on the infrared image based on the surface temperature of the target chip.
10. The appearance inspection equipment for chip manufacturing according to claim 9, characterized in that, The operating temperature range of the acquisition component is set to [33 degrees Celsius - 36 degrees Celsius].