Sapphire Substrate Cleaning Data Processing System and Method Based on Image Recognition

Abstract

The present invention discloses a sapphire substrate cleaning data processing system and method based on image recognition, belonging to the field of semiconductor processing. The present invention obtains the data of the change in the dirt on the surface of the sapphire substrate during the cleaning process and imports it into the dirt cleaning anomaly analysis strategy for dirt cleaning anomaly analysis. The obtained defect cleaning anomaly analysis results and dirt cleaning anomaly analysis results are imported into the cleaning water pressure analysis model for cleaning water pressure analysis. The cleaning water pressure data during the cleaning process is adjusted according to the analyzed cleaning water pressure. By means of image processing, the dirt and defects on the cleaned sapphire surface are accurately distinguished, and the cleaning water pressure is accurately adjusted according to the defect change data and dirt cleaning change data during the cleaning process, improving the cleaning efficiency while reducing the damage to the sapphire substrate during the cleaning process.

Description

Technical Field

[0001] The present invention belongs to the field of semiconductor processing, and specifically relates to a sapphire substrate cleaning data processing system and method based on image recognition. Background Art

[0002] A sapphire substrate is an important material used in semiconductor manufacturing. It is mainly made of synthetic sapphire crystals. In the fields of LEDs (light-emitting diodes) and optoelectronics, the sapphire substrate is widely used due to its unique physical and chemical properties. The sapphire substrate is a substrate for epitaxial growth of semiconductor materials and is usually used as the substrate of an LED chip. It is made of high-purity aluminum oxide crystals through specific processing techniques, and has extremely high hardness and thermal stability. The manufacture of the sapphire substrate involves melting high-purity aluminum oxide and then growing crystals through specific methods (such as the Czochralski method or the edge-defined film-fed growth method), and then cutting, polishing, and cleaning these crystals to obtain the required size and surface quality. It plays a key role in LED manufacturing and other optoelectronic applications with its unique physical and chemical properties;

[0003] After the production of the sapphire substrate, a cleaning device is needed to clean the dirt on the sapphire substrate. However, during the cleaning process, it is impossible to accurately distinguish the dirt and defects on the surface of the cleaned sapphire, and it is impossible to accurately adjust the cleaning water pressure according to the defect change data and dirt cleaning change data during the cleaning process, resulting in a reduction in cleaning efficiency and damage to the sapphire substrate during the cleaning process. Most of the existing technologies have the above problems;

[0004] For example, in the Chinese patent with the application publication number CN105931949A, a single-wafer cleaning method for a patterned sapphire substrate rework wafer is disclosed, which solves the problem that the existing multi-wafer cleaning method is difficult to remove the photoresist because the photoresist remaining on the surface of the patterned sapphire wafer has stable chemical properties, strong surface adsorption ability, and is not easily dissolved by ordinary acids and alkalis. The specific steps of this method are as follows: ⑴ Select a patterned sapphire substrate rework wafer and perform spin cleaning with a high-temperature degluer; ⑵ Perform spin cleaning of the wafer with isopropyl alcohol; ⑶ Rinse the wafer with deionized water; ⑷ Brush the wafer; ⑸ Spin dry. The steps ⑴ to ⑸ are all carried out under the cleanroom standard with a dust particle content reaching class 100. This invention avoids the use of corrosive chemical substances such as strong acids and strong alkalis, reduces the damage to front-line employees and equipment pipelines, and multiple single-wafer cycles of cleaning ensure the cleaning quality of each sapphire substrate wafer;

[0005] The above patents all have the problems proposed in this background art. To solve the problems proposed in this background art, the present application designs a sapphire substrate cleaning data processing system and method based on image recognition. Summary of the Invention

[0006] In view of the deficiencies of the prior art, the present invention proposes a sapphire substrate cleaning data processing system and method based on image recognition. The present invention acquires image data of the sapphire substrate during the cleaning process, and at the same time acquires cleaning water pressure data during the cleaning process, distinguishes surface defects and dirt of the sapphire, acquires defect change data of the sapphire substrate during the cleaning process and imports it into the defect cleaning anomaly analysis strategy for defect cleaning anomaly analysis, acquires surface dirt change data of the sapphire substrate during the cleaning process and imports it into the dirt cleaning anomaly analysis strategy for dirt cleaning anomaly analysis, imports the obtained defect cleaning anomaly analysis result and dirt cleaning anomaly analysis result into the cleaning water pressure analysis model for cleaning water pressure analysis, adjusts the cleaning water pressure data during the cleaning process according to the analyzed cleaning water pressure, accurately distinguishes dirt and defects on the surface of the cleaned sapphire through image processing means, accurately adjusts the cleaning water pressure according to the defect change data and dirt cleaning change data during the cleaning process, and improves the cleaning efficiency while reducing the damage of the sapphire substrate during the cleaning process.

[0007] To achieve the above object, the present invention provides the following technical solutions: A sapphire substrate cleaning data processing method based on image recognition, which includes the following specific steps:

[0008] Acquire image data of the sapphire substrate during the cleaning process, and at the same time acquire cleaning water pressure data during the cleaning process, and distinguish surface defects and dirt of the sapphire;

[0009] Acquire defect change data of the sapphire substrate during the cleaning process and import it into the defect cleaning anomaly analysis strategy for defect cleaning anomaly analysis;

[0010] Acquire surface dirt change data of the sapphire substrate during the cleaning process and import it into the dirt cleaning anomaly analysis strategy for dirt cleaning anomaly analysis;

[0011] Import the obtained defect cleaning anomaly analysis result and dirt cleaning anomaly analysis result into the cleaning water pressure analysis model for cleaning water pressure analysis;

[0012] Adjust the cleaning water pressure data during the cleaning process according to the analyzed cleaning water pressure.

[0013] It should be noted here that as a preferred technical solution of the sapphire substrate cleaning data processing method based on image recognition, the specific steps of acquiring image data of the sapphire substrate during the cleaning process, and at the same time acquiring cleaning water pressure data during the cleaning process, and distinguishing surface defects and dirt of the sapphire are as follows:

[0014] S11. Obtain the set water pressure data during the cleaning process, obtain the image data of the sapphire substrate in real time during the cleaning process, import it into the image processing software to obtain the pixel value data of each pixel point in the image, and at the same time obtain the average pixel value data of the corresponding area of the sapphire substrate. Obtain the pixel points whose absolute value of the difference between the pixel value of the pixel point in the image and the average pixel value of the corresponding area is greater than or equal to the difference threshold, and set them as abnormal pixel points. Set the pixel value gradient, obtain the contour formed by the abnormal pixel points in the same pixel value gradient, and set it as the judgment contour. Obtain the contour data and pixel point pixel value data of several judgment contours;

[0015] S12. Obtain the contour data and pixel point pixel value data of the judgment contour, and at the same time obtain the contours of historical defects and dirt, and the average pixel point pixel value data, and substitute them into the judgment value calculation formula to calculate the judgment value of the judgment contour. Among them, the judgment value calculation formula is: , where a is the contour similarity ratio coefficient, c() is the area of the contour in the parentheses, Si is the i-th contour in the set of historical defect and dirt contours, S is the contour of the judgment contour, xi is the average pixel value of the contour in the set of historical defect and dirt contours, and x is the average pixel value data of the judgment contour;

[0016] S13. Obtain the types of historical defects and dirt corresponding to the maximum judgment value as the type of the judgment contour, obtain the type data of all judgment contours, and classify the judgment contours into defects and dirt.

[0017] It should be noted here that as an optimal technical solution of the sapphire substrate cleaning data processing method based on image recognition, the obtaining of the sapphire substrate defect change data during the cleaning process and importing it into the defect cleaning anomaly analysis strategy for defect cleaning anomaly analysis includes the following specific steps:

[0018] S21. Obtain the change image of the defect during the cleaning process and the image of the defect at the starting moment, and obtain the area of the defect and the change data of the pixel point pixel value in the image;

[0019] S22. Obtain the image of the defect at the starting moment and import it into the defect anomaly value calculation formula to calculate the starting defect anomaly value. Among them, the defect anomaly value calculation formula is: , where n is the number of defects, mz is the number of pixel points of the z-th defect, sz is the area of the z-th defect, S is the area of the sapphire substrate, c is the distance from the standard value, czj is the distance from the j-th pixel point of the z-th defect to the center point of the defect, where the center point of the defect is the point with the smallest sum of distances to other pixel points of the corresponding defect, kzj is the pixel value of the j-th pixel point of the z-th defect, and k is the average pixel value data of the corresponding area of the sapphire substrate;

[0020] S23. Obtain the images of the defects in real time during the cleaning process and import them into the defect outlier calculation formula to calculate the cleaning defect outlier value;

[0021] S24. Substitute the obtained starting defect outlier value and cleaning defect outlier value into the defect cleaning anomaly analysis value calculation formula to calculate the defect cleaning anomaly analysis value. The defect cleaning anomaly analysis value calculation formula is as follows: , where xm is the cleaning defect outlier value, T is the time standard value, and tx is the cleaning time.

[0022] Here, it should be noted that as an optimal technical solution of the sapphire substrate cleaning data processing method based on image recognition, the obtaining of the surface fouling change data of the sapphire substrate during the cleaning process and importing it into the fouling cleaning anomaly analysis strategy for fouling cleaning anomaly analysis includes the following specific steps:

[0023] S31. Obtain the fouling change image during the cleaning process and the fouling image at the starting moment, and obtain the fouling area and pixel value change data of the pixels in the image;

[0024] S32. Substitute the obtained fouling image at the starting moment into the fouling analysis value calculation formula to calculate the starting moment fouling analysis value. The fouling analysis value calculation formula is as follows: , where P is the number of fouling contours at the starting moment, mp is the number of pixels of the p-th fouling contour, and Qpt is the pixel value of the t-th pixel of the p-th fouling contour;

[0025] S33. Import the obtained real-time fouling image into the fouling analysis value calculation formula to calculate the real-time fouling analysis value;

[0026] S34. Substitute the calculated real-time fouling analysis value and the starting moment fouling analysis value into the fouling cleaning value calculation formula to calculate the fouling cleaning value. The fouling cleaning value calculation formula is as follows: , where w is the fouling analysis safety value, is the cleaning difference ratio coefficient, and wx is the real-time fouling analysis value.

[0027] Here, it should be noted that as an optimal technical solution of the sapphire substrate cleaning data processing method based on image recognition, the importing of the obtained defect cleaning anomaly analysis result and fouling cleaning anomaly analysis result into the cleaning water pressure analysis model for cleaning water pressure analysis includes the following specific contents:

[0028] S41. Determine whether the fouling analysis value is less than the fouling analysis safety value. If the fouling analysis value is less than or equal to the fouling analysis safety value, it indicates that the cleaning is complete; if the fouling analysis value is greater than the fouling analysis safety value, it indicates that further cleaning is required, and proceed to S42; S42. Obtain the calculated defect cleaning anomaly analysis value, fouling cleaning value, and set water pressure data, and substitute them into the cleaning water pressure calculation formula to calculate the cleaning water pressure. The cleaning water pressure calculation formula is: , where fm is the set water pressure data, ln() is the natural logarithm with the natural constant e as the base, and C is the compensation coefficient.

[0029] A sapphire substrate cleaning data processing system based on image recognition, which is implemented based on the above-mentioned sapphire substrate cleaning data processing method based on image recognition. It specifically includes an image acquisition module, an image discrimination module, a defect cleaning anomaly analysis module, a fouling cleaning anomaly analysis module, a cleaning water pressure analysis module, and a water pressure adjustment module. Among them, the image acquisition module is used to acquire the image data of the sapphire substrate during the cleaning process, and at the same time acquire the cleaning water pressure data during the cleaning process;

[0030] The image discrimination module is used to distinguish the surface defects and fouling of the sapphire through image processing means;

[0031] The defect cleaning anomaly analysis module is used to obtain the sapphire substrate defect change data during the cleaning process and import it into the defect cleaning anomaly analysis strategy for defect cleaning anomaly analysis;

[0032] The fouling cleaning anomaly analysis module is used to obtain the surface fouling change data of the sapphire substrate during the cleaning process and import it into the fouling cleaning anomaly analysis strategy for fouling cleaning anomaly analysis;

[0033] The cleaning water pressure analysis module is used to import the obtained defect cleaning anomaly analysis result and fouling cleaning anomaly analysis result into the cleaning water pressure analysis model for cleaning water pressure analysis;

[0034] The water pressure adjustment module is used to adjust the cleaning water pressure data during the cleaning process according to the analyzed cleaning water pressure.

[0035] An electronic device includes: a processor and a memory. Among them, the memory stores a computer program that can be called by the processor;

[0036] The processor executes the above-mentioned sapphire substrate cleaning data processing method based on image recognition by calling the computer program stored in the memory.

[0037] A computer-readable storage medium stores instructions. When the instructions run on a computer, the computer is made to execute the sapphire substrate cleaning data processing method based on image recognition as described above.

[0038] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0039] The present invention acquires image data of a sapphire substrate during the cleaning process, and at the same time acquires cleaning water pressure data during the cleaning process, differentiates surface defects and dirt of the sapphire, acquires defect change data of the sapphire substrate during the cleaning process and imports it into a defect cleaning anomaly analysis strategy for defect cleaning anomaly analysis, acquires surface dirt change data of the sapphire substrate during the cleaning process and imports it into a dirt cleaning anomaly analysis strategy for dirt cleaning anomaly analysis, imports the obtained defect cleaning anomaly analysis result and dirt cleaning anomaly analysis result into a cleaning water pressure analysis model for cleaning water pressure analysis, adjusts the cleaning water pressure data during the cleaning process according to the analyzed cleaning water pressure, accurately distinguishes dirt and defects on the surface of the cleaned sapphire through image processing means, accurately adjusts the cleaning water pressure according to the defect change data and dirt cleaning change data during the cleaning process, and reduces the damage of the sapphire substrate during the cleaning process while improving the cleaning efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic diagram of the overall process of the sapphire substrate cleaning data processing method based on image recognition according to the present invention;

[0041] Figure 2 It is a schematic diagram of step S2 of the sapphire substrate cleaning data processing method based on image recognition according to the present invention;

[0042] Figure 3 It is a schematic diagram of the step of differentiating defects and dirt of the sapphire substrate cleaning data processing method based on image recognition according to the present invention;

[0043] Figure 4 It is a schematic diagram of the sapphire substrate cleaning data processing system based on image recognition according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The description of at least one exemplary embodiment below is actually only illustrative and in no way limits the present application and its application or use.

[0045] Embodiment 1

[0046] To solve the technical problems raised in the background art: In the prior art, it is impossible to accurately distinguish the dirt and defects on the surface of the cleaned sapphire during the cleaning process, and it is impossible to accurately adjust the cleaning water pressure according to the defect change data and dirt cleaning change data during the cleaning process, resulting in a reduction in cleaning efficiency and damage to the sapphire substrate during the cleaning process; The present invention provides a preferred embodiment: As Figure 1 - Figure 2 shown, a method for processing sapphire substrate cleaning data based on image recognition, which includes the following specific steps:

[0047] S1. Obtain the image data of the sapphire substrate during the cleaning process, and at the same time obtain the cleaning water pressure data during the cleaning process to distinguish the surface defects and dirt of the sapphire;

[0048] In this embodiment, the specific steps for obtaining the image data of the sapphire substrate during the cleaning process and at the same time obtaining the cleaning water pressure data during the cleaning process to distinguish the surface defects and dirt of the sapphire are as follows:

[0049] S11. Obtain the set water pressure data during the cleaning process, obtain the image data of the sapphire substrate during the cleaning process in real time, import it into the image processing software to obtain the pixel value data of each pixel point in the image, and at the same time obtain the average pixel value data of the corresponding area of the sapphire substrate. Obtain the pixel points whose absolute value of the difference between the pixel value of the pixel point in the image and the average pixel value of the corresponding area is greater than or equal to the difference threshold, and set them as abnormal pixel points. Set the pixel value gradient, and obtain the contour formed by the abnormal pixel points in the same pixel value gradient, and set it as the judgment contour. Obtain the contour data and pixel point pixel value data of several judgment contours;

[0050] S12. Obtain the contour data and pixel point pixel value data of the judgment contour, and at the same time obtain the contours and average pixel point pixel value data that have been historically identified as defects and dirt, and substitute them into the judgment value calculation formula to calculate the judgment value of the judgment contour. Among them, the judgment value calculation formula is: , where a is the contour similarity ratio coefficient, c() is the area of the contour in the parentheses, Si is the i-th contour in the set of historical defect and dirt contours, S is the contour of the judgment contour, xi is the average pixel value of the contour in the set of historical defect and dirt contours, and x is the average pixel value data of the judgment contour;

[0051] S13. Obtain the types of historical defects and dirt corresponding to the maximum judgment value as the types of the judgment contour, obtain the type data of all judgment contours, and classify the judgment contours into defects and dirt;

[0052] In this step, image processing and image segmentation means are used to accurately identify and classify defects and dirt;

[0053] S2. Obtain the defect change data during the cleaning process of the sapphire substrate and import it into the defect cleaning anomaly analysis strategy for defect cleaning anomaly analysis;

[0054] In this embodiment, obtaining the defect change data during the cleaning process of the sapphire substrate and importing it into the defect cleaning anomaly analysis strategy for defect cleaning anomaly analysis includes the following specific steps:

[0055] S21. Obtain the change image of the defects during the cleaning process and the image of the defects at the starting moment, and obtain the area of the defects and the change data of the pixel values of the pixel points in the image;

[0056] S22. Obtain the image of the defects at the starting moment and import it into the defect anomaly value calculation formula to calculate the starting defect anomaly value. Among them, the defect anomaly value calculation formula is: , where n is the number of defects, mz is the number of pixel points of the z-th defect, sz is the area of the z-th defect, S is the area of the sapphire substrate, c is the distance from the standard value, czj is the distance from the j-th pixel point of the z-th defect to the center point of the defect, where the center point of the defect is the point with the smallest sum of distances to other pixel points of the corresponding defect, kzj is the pixel value of the j-th pixel point of the z-th defect, and k is the average pixel value data of the corresponding area of the sapphire substrate;

[0057] S23. Obtain the real-time image of the defects during the cleaning process and import it into the defect anomaly value calculation formula to calculate the cleaning defect anomaly value;

[0058] S24. Substitute the obtained starting defect anomaly value and cleaning defect anomaly value into the defect cleaning anomaly analysis value calculation formula to calculate the defect cleaning anomaly analysis value. Among them, the defect cleaning anomaly analysis value calculation formula is: , where xm is the cleaning defect anomaly value, T is the time standard value, and tx is the cleaning time;

[0059] In this step, by analyzing the defect change data during the cleaning process, the damage to the sapphire substrate during the cleaning process is accurately evaluated;

[0060] S3. Obtain the surface dirt change data of the sapphire substrate during the cleaning process and import it into the dirt cleaning anomaly analysis strategy for dirt cleaning anomaly analysis;

[0061] In this embodiment, obtaining the surface dirt change data of the sapphire substrate during the cleaning process and importing it into the dirt cleaning anomaly analysis strategy for dirt cleaning anomaly analysis includes the following specific steps:

[0062] S31. Obtain the change image of the dirt during the cleaning process and the image of the dirt at the starting moment, and obtain the area of the dirt and the change data of the pixel values of the pixel points in the image;

[0063] S32. Substitute the obtained image of dirt at the starting moment into the dirt analysis value calculation formula to calculate the dirt analysis value at the starting moment. The dirt analysis value calculation formula is as follows: , where P is the number of dirt contours at the starting moment, mp is the number of pixel points of the p-th dirt contour, and Qpt is the pixel value of the t-th pixel point of the p-th dirt contour;

[0064] S33. Import the obtained real-time image of dirt into the dirt analysis value calculation formula to calculate the real-time dirt analysis value;

[0065] S34. Substitute the calculated real-time dirt analysis value and the dirt analysis value at the starting moment into the dirt cleaning value calculation formula to calculate the dirt cleaning value. The dirt cleaning value calculation formula is as follows: , where w is the dirt analysis safety value, is the cleaning difference ratio coefficient, and wx is the real-time dirt analysis value; the dirt analysis safety value here is set by manual experience and is set according to the required standard of the dirt degree on the substrate surface;

[0066] S4. Import the obtained defect cleaning anomaly analysis result and dirt cleaning anomaly analysis result into the cleaning water pressure analysis model to analyze the cleaning water pressure;

[0067] In this embodiment, importing the obtained defect cleaning anomaly analysis result and dirt cleaning anomaly analysis result into the cleaning water pressure analysis model to analyze the cleaning water pressure includes the following specific contents:

[0068] S41. Determine whether the dirt analysis value is less than the dirt analysis safety value. If the dirt analysis value is less than or equal to the dirt analysis safety value, it means that the cleaning is complete; if the dirt analysis value is greater than the dirt analysis safety value, it means that further cleaning is required, and go to S42; S42. Obtain the calculated defect cleaning anomaly analysis value, dirt cleaning value, and set water pressure data, and substitute them into the cleaning water pressure calculation formula to calculate the cleaning water pressure. The cleaning water pressure calculation formula is as follows: , where fm is the set water pressure data, ln() is the natural logarithm with the natural constant e as the base, and C is the compensation coefficient;

[0069] S5. Adjust the cleaning water pressure data during the cleaning process according to the analyzed cleaning water pressure.

[0070] It should be noted here that the compensation coefficient and the cleaning difference ratio coefficient are obtained based on the experience of those skilled in the art. The preferred way to obtain the values of the compensation coefficient and the cleaning difference ratio coefficient is as follows: Obtain a number of groups of image data of the sapphire substrate during the cleaning process and the cleaning water pressure data during the cleaning process. Substitute the image data of the sapphire substrate during the cleaning process and the cleaning water pressure data during the cleaning process into the cleaning water pressure calculation formula to calculate the cleaning water pressure. Hire experts to score the sapphire substrate after cleaning. Substitute the scoring results and the cleaning water pressure calculation results into the fitting software to output the values of the compensation coefficient and the cleaning difference ratio coefficient that meet the highest scoring sorting accuracy rate.

[0071] Here, the benefits of this embodiment need to be specifically described. The benefits of this embodiment compared with the prior art are as follows: Obtain the image data of the sapphire substrate during the cleaning process, and at the same time obtain the cleaning water pressure data during the cleaning process, distinguish the surface defects and dirt of the sapphire, obtain the defect change data of the sapphire substrate during the cleaning process and import it into the defect cleaning anomaly analysis strategy for defect cleaning anomaly analysis, obtain the surface dirt change data of the sapphire substrate during the cleaning process and import it into the dirt cleaning anomaly analysis strategy for dirt cleaning anomaly analysis, import the obtained defect cleaning anomaly analysis results and dirt cleaning anomaly analysis results into the cleaning water pressure analysis model for cleaning water pressure analysis, adjust the cleaning water pressure data during the cleaning process according to the analyzed cleaning water pressure, accurately distinguish the dirt and defects on the surface of the cleaned sapphire through image processing means, and accurately adjust the cleaning water pressure according to the defect change data and dirt cleaning change data during the cleaning process, which improves the cleaning efficiency and reduces the damage of the sapphire substrate during the cleaning process.

[0072] Embodiment 2

[0073] As Figure 4 shown, the sapphire substrate cleaning data processing system based on image recognition is implemented based on the above-mentioned sapphire substrate cleaning data processing method based on image recognition. It specifically includes an image acquisition module, an image discrimination module, a defect cleaning anomaly analysis module, a dirt cleaning anomaly analysis module, a cleaning water pressure analysis module, and a water pressure adjustment module. Among them, the image acquisition module is used to obtain the image data of the sapphire substrate during the cleaning process and at the same time obtain the cleaning water pressure data during the cleaning process;

[0074] The image discrimination module is used to distinguish the surface defects and dirt of the sapphire through image processing means;

[0075] The defect cleaning anomaly analysis module is used to obtain the defect change data of the sapphire substrate during the cleaning process and import it into the defect cleaning anomaly analysis strategy for defect cleaning anomaly analysis;

[0076] The dirty cleaning anomaly analysis module is used to obtain the data on the change of dirt on the surface of the sapphire substrate during the cleaning process and import it into the dirty cleaning anomaly analysis strategy for dirty cleaning anomaly analysis;

[0077] The cleaning water pressure analysis module is used to import the obtained defect cleaning anomaly analysis results and dirty cleaning anomaly analysis results into the cleaning water pressure analysis model for the analysis of the cleaning water pressure;

[0078] The water pressure adjustment module is used to adjust the cleaning water pressure data during the cleaning process according to the analyzed cleaning water pressure;

[0079] The connection lines in the attached drawings represent the signal transmission relationships of each module, and at the same time, each module is controlled by the server to run.

[0080] Embodiment 3

[0081] This embodiment provides an electronic device, including: a processor and a memory, wherein, a computer program that can be called by the processor is stored in the memory;

[0082] The processor executes the above-mentioned sapphire substrate cleaning data processing method based on image recognition by calling the computer program stored in the memory.

[0083] This electronic device may have relatively large differences due to different configurations or performances, and can include one or more processors and one or more memories. Among them, at least one computer program is stored in the memory, and this computer program is loaded and executed by the processor to implement the sapphire substrate cleaning data processing method based on image recognition provided by the above method embodiment. This electronic device can also include other components for realizing the functions of the device. For example, this electronic device can also have components such as wired or wireless network interfaces and input / output interfaces for data input and output. This embodiment will not be elaborated here.

[0084] Embodiment 4

[0085] This embodiment proposes a computer-readable storage medium, on which a rewritable computer program is stored;

[0086] When the computer program runs on a computer device, it causes the computer device to execute the above-mentioned sapphire substrate cleaning data processing method based on image recognition.

[0087] For example, the computer-readable storage medium can be a read-only memory, a random access memory, a compact disc read-only memory, magnetic tape, a floppy disk, and an optical data storage device, etc.

[0088] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions according to the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired network or / and a wireless network. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

Claims

1. A sapphire substrate cleaning data processing method based on image recognition, characterized in that: It includes the following specific steps: Acquire image data of the sapphire substrate during the cleaning process, and simultaneously acquire cleaning water pressure data during the cleaning process to distinguish defects and contamination on the sapphire surface; Obtain the sapphire substrate defect change data during the cleaning process and import it into the defect cleaning anomaly analysis strategy to perform defect cleaning anomaly analysis; Obtain the contamination change data of the sapphire substrate surface during the cleaning process and import it into the contamination cleaning abnormality analysis strategy to perform contamination cleaning abnormality analysis; Importing the acquired defect cleaning abnormality analysis results and contamination cleaning abnormality analysis results into the cleaning water pressure analysis model to analyze the cleaning water pressure; The cleaning water pressure data during the cleaning process is adjusted according to the cleaning water pressure obtained by analysis; the image data of the sapphire substrate during the cleaning process is obtained, and the cleaning water pressure data during the cleaning process is obtained at the same time, and the specific steps of distinguishing the surface defects and contamination of the sapphire are: S11, obtaining the set water pressure data during the cleaning process, obtaining the image data of the sapphire substrate during the cleaning process in real time, importing the image processing software to obtain the pixel value data of each pixel in the image, and obtaining the average pixel value data of the corresponding area of ​​the sapphire substrate at the same time, obtaining the pixel points whose absolute value of the difference between the pixel value of the pixel point in the image and the average pixel value of the corresponding area is greater than or equal to the difference threshold, setting them as abnormal pixel points, setting the pixel value gradient, obtaining the contour formed by the abnormal pixel points in the same pixel value gradient, setting it as the judgment contour, and obtaining the contour data of several judgment contours and the pixel value data of the pixel points; S12, obtaining contour data and pixel value data of the judgment contour, and simultaneously obtaining contours and average pixel value data that have been historically identified as defects and contamination, and substituting them into the judgment value calculation formula to calculate the judgment value of the judgment contour, wherein the judgment value calculation formula is: , where a is the contour similarity ratio coefficient, c() is the area of ​​the contour in the brackets, Si is the i-th contour in the set of historical defect and dirty contours, S is the contour of the judgment contour, xi is the average pixel value of the contour in the set of historical defect and dirty contours, and x is the average pixel value data of the judgment contour; S13, obtaining the types of historical defects and contamination corresponding to the maximum judgment value as the types of judgment contours, obtaining type data of all judgment contours, and classifying the judgment contours into defects and contamination.

2. The sapphire substrate cleaning data processing method based on image recognition according to claim 1, characterized in that: The method of obtaining the sapphire substrate defect change data during the cleaning process and importing it into the defect cleaning abnormality analysis strategy for defect cleaning abnormality analysis includes the following specific steps: S21, obtaining an image of defect changes during the cleaning process and an image of the defect at the start time, and obtaining defect area and pixel value change data of the pixel points in the image; S22, obtaining an image of the defect at the starting time and importing it into a defect abnormal value calculation formula to calculate the starting defect abnormal value, wherein the defect abnormal value calculation formula is: , where n is the number of defects, mz is the number of pixels of the z-th defect, sz is the area of ​​the z-th defect, S is the area of ​​the sapphire substrate, c is the distance standard value, czj is the distance from the j-th pixel of the z-th defect to the center of the defect, where the center of the defect is the point with the smallest distance from the other pixels of the corresponding defect, kzj is the pixel value of the j-th pixel of the z-th defect, and k is the average pixel value data of the corresponding area of ​​the sapphire substrate; S23, obtaining real-time defect images during the cleaning process and importing them into a defect abnormality value calculation formula to calculate a cleaning defect abnormality value; S24, substituting the obtained initial defect abnormal value and cleaning defect abnormal value into the defect cleaning abnormal analysis value calculation formula to calculate the defect cleaning abnormal analysis value, wherein the defect cleaning abnormal analysis value calculation formula is: , where xm is the cleaning defect abnormal value, T is the time standard value, and tx is the cleaning time.

3. The sapphire substrate cleaning data processing method based on image recognition according to claim 2, characterized in that: The method of obtaining the sapphire substrate surface contamination change data during the cleaning process and importing it into the contamination cleaning abnormality analysis strategy to perform contamination cleaning abnormality analysis includes the following specific steps: S31, obtaining an image of the change of dirt during the cleaning process and an image of the dirt at the start time, and obtaining data on the change of the dirt area and pixel values ​​of the pixels in the image; S32, substituting the obtained pollution image at the starting time into the pollution analysis value calculation formula to calculate the pollution analysis value at the starting time, wherein the pollution analysis value calculation formula is: , where P is the number of dirty contours at the starting moment, mp is the number of pixels of the pth dirty contour, and Qpt is the pixel value of the tth pixel of the pth dirty contour.

4. The sapphire substrate cleaning data processing method based on image recognition according to claim 3, characterized in that: The method of obtaining the sapphire substrate surface contamination change data during the cleaning process and importing it into the contamination cleaning abnormality analysis strategy for performing contamination cleaning abnormality analysis also includes the following specific steps: S33, the obtained real-time pollution image is imported into the pollution analysis value calculation formula to calculate the real-time pollution analysis value; S34, substituting the calculated real-time pollution analysis value and the starting time pollution analysis value into the pollution cleaning value calculation formula to calculate the pollution cleaning value, wherein the pollution cleaning value calculation formula is: , where w is the pollution analysis safety value, is the cleaning difference ratio, and wx is the real-time dirt analysis value.

5. The sapphire substrate cleaning data processing method based on image recognition according to claim 4, characterized in that: The method of importing the obtained defect cleaning abnormality analysis results and the dirty cleaning abnormality analysis results into the cleaning water pressure analysis model to analyze the cleaning water pressure includes the following specific contents: S41, determine whether the turbidity analysis value is less than the turbidity analysis safety value. If the turbidity analysis value is less than or equal to the turbidity analysis safety value, it means that the cleaning is complete; if the turbidity analysis value is greater than the turbidity analysis safety value, it means that cleaning needs to be continued, and S42 is performed; S42, obtain the calculated defect cleaning abnormality analysis value, the turbidity cleaning value and the set water pressure data, and substitute them into the cleaning water pressure calculation formula to calculate the cleaning water pressure, wherein the cleaning water pressure calculation formula is: , where fm is the set water pressure data, ln() is the logarithm with the natural constant e as the base, and C is the compensation coefficient.

6. A sapphire substrate cleaning data processing system based on image recognition, which is implemented based on the sapphire substrate cleaning data processing method based on image recognition as claimed in any one of claims 1 to 5, characterized in that: It specifically includes an image acquisition module, an image differentiation module, a defect cleaning abnormality analysis module, a contamination cleaning abnormality analysis module, a cleaning water pressure analysis module and a water pressure adjustment module, wherein the image acquisition module is used to acquire image data of the sapphire substrate during the cleaning process and simultaneously acquire cleaning water pressure data during the cleaning process; The image distinguishing module is used to distinguish sapphire surface defects and contamination by image processing means, wherein the specific steps of distinguishing sapphire surface defects and contamination are: S11, obtaining the set water pressure data during the cleaning process, obtaining the image data of the sapphire substrate during the cleaning process in real time, importing the image processing software to obtain the pixel value data of each pixel in the image, and obtaining the average pixel value data of the corresponding area of ​​the sapphire substrate at the same time, obtaining the pixel points whose absolute value of the difference between the pixel value of the pixel point in the image and the average pixel value of the corresponding area is greater than or equal to the difference threshold, setting them as abnormal pixel points, setting the pixel value gradient, obtaining the contour formed by the abnormal pixel points in the same pixel value gradient, setting it as the judgment contour, and obtaining the contour data of several judgment contours and the pixel value data of the pixel points; S12, obtaining contour data and pixel value data of the judgment contour, and simultaneously obtaining contours and average pixel value data that have been historically identified as defects and contamination, and substituting them into the judgment value calculation formula to calculate the judgment value of the judgment contour, wherein the judgment value calculation formula is: , where a is the contour similarity ratio coefficient, c() is the area of ​​the contour in the brackets, Si is the i-th contour in the set of historical defect and dirty contours, S is the contour of the judgment contour, xi is the average pixel value of the contour in the set of historical defect and dirty contours, and x is the average pixel value data of the judgment contour; S13, obtaining the types of historical defects and contamination corresponding to the maximum judgment value as the types of judgment contours, obtaining type data of all judgment contours, and classifying the judgment contours into defects and contamination; The defect cleaning anomaly analysis module is used to obtain the sapphire substrate defect change data during the cleaning process and import it into the defect cleaning anomaly analysis strategy to perform defect cleaning anomaly analysis; The dirt cleaning abnormality analysis module is used to obtain the dirt change data of the sapphire substrate surface during the cleaning process and import it into the dirt cleaning abnormality analysis strategy to perform dirt cleaning abnormality analysis; The cleaning water pressure analysis module is used to import the acquired defect cleaning abnormality analysis results and the contamination cleaning abnormality analysis results into the cleaning water pressure analysis model to analyze the cleaning water pressure; The water pressure regulating module is used to regulate the cleaning water pressure data during the cleaning process according to the cleaning water pressure obtained by analysis.

7. An electronic device comprising: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; It is characterized in that the processor executes the sapphire substrate cleaning data processing method based on image recognition as described in any one of claims 1 to 5 by calling the computer program stored in the memory.

8. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are executed on a computer, the computer is caused to execute the sapphire substrate cleaning data processing method based on image recognition as described in any one of claims 1 to 5.

Images