Thermal spraying coating quality detection system and method based on machine vision

Abstract

The invention relates to the technical field of image data processing, in particular to a thermal spraying coating quality detection system and method based on machine vision, and the method comprises the steps: collecting image index parameters; analyzing the image index characterization value based on the image index parameter; based on a comparison result of the image index characterization value and a preset image index characterization threshold value, determining whether the target sample coating image meets a standard or not; in response to the condition that the coating image of the target sample meets the standard, collecting performance index parameters of the target sample in a historical period; analyzing a performance index characterization value based on the performance index parameter; determining whether the coating quality of the target sample meets a standard or not based on a difference result between the performance index characterization value and a preset performance index characterization threshold value; and in response to the condition that the coating quality of the target sample does not meet the standard, determining a corresponding processing strategy based on a difference value between the performance index characterization value and a preset performance index characterization threshold value. According to the invention, the quality detection efficiency of machine vision is improved through self-adaptive regulation and control.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image data processing, and particularly relates to a thermal spraying coating quality detection system and method based on machine vision. BACKGROUND

[0002] Machine vision technology shows significant technical adaptability and application advantages in the remanufacturing and repair of hydraulic cylinder piston rods. As a standard cylinder, the regular geometric characteristics of the piston rod create ideal conditions for high-precision positioning, three-dimensional scanning and global measurement of the vision system, effectively reducing the complexity and difficulty of the detection scene. At present, mature repair technologies represented by electric arc spraying and plasma spraying ensure the stability of the coating forming process, providing a solid foundation for the vision system to establish a standardized and reusable detection model. In the quality evaluation link, core indicators such as coating thickness, surface roughness and macroscopic defects can be converted into high-dimensional image features, thereby highly matching the quantitative analysis capability of machine vision. By introducing an integrated vision detection system, consistent and efficient judgment of repair quality can be achieved, significantly improving the reliability, success rate and comprehensive benefits of the remanufacturing process. Not only can it replace manual labor to achieve non-destructive and efficient online detection, but also can optimize the spraying process parameters through massive data iteration, ultimately building an intelligent manufacturing closed loop from precise detection to process optimization.

[0003] Chinese patent publication No. CN117871410A discloses a machine vision automatic detection system and method for coating appearance quality, which includes the following steps: S1: checking the detection device and dividing the products to be detected into different detection areas; S2: collecting data of each detection area to complete the detection process. The present application can avoid the problems of human eye unreachability and easy missed detection during large-size coating product appearance detection, ensuring 100% detection of large-size products in all positions, and can be widely applied in the field of coating detection. The provided method is easy to implement and does not require manual detection with the naked eye, greatly reducing the cost of manpower and resources.

[0004] Chinese Patent Publication No. CN120182247A discloses a machine vision-based method and system for detecting the coating quality of injection molded parts. The method includes the following steps: determining pixels in deep bright areas and edge pixels of deep bright areas based on the difference between the grayscale value of a pixel in a surface grayscale image and the maximum grayscale value, as well as the average grayscale difference between the pixel and its neighboring areas at different scales; obtaining pixels in light bright areas based on the probability that the remaining pixels are in deep bright areas, the difference between the brightness value and the standard value, and the chromaticity distance between the remaining pixels and the remaining pixels corresponding to the standard value; and training the pixels in deep bright areas and their corresponding edge pixels, along with the pixels in light bright areas and their corresponding edge pixels, in the surface image of the injection molded part after enhancement, to achieve coating quality detection of the injection molded part, effectively improving the accuracy and efficiency of coating quality detection.

[0005] Therefore, it is evident that the existing technology has the following problems: In existing technologies, the detection dimensions of machine vision are single and isolated, making it impossible to establish a correlation model between image features and quality parameters. This results in machine vision being unable to quantify the quality detection process in real time, leading to a decrease in quality detection efficiency. Summary of the Invention

[0006] To address this issue, the present invention provides a machine vision-based thermal spray coating quality inspection system and method to overcome the problem in the prior art where the machine vision detection dimensions are single and isolated, making it impossible to establish a correlation model between image features and quality parameters, thus preventing the machine vision from quantifying the quality inspection process in real time and causing a decrease in quality inspection efficiency.

[0007] To achieve the aforementioned objective, the present invention provides a machine vision-based method for inspecting the quality of thermal spray coatings, comprising: Collect image index parameters of the target sample within a historical period; Analyze the image index characterization values ​​based on the aforementioned image index parameters; Based on the comparison between the image index characterization value and the predetermined image index characterization threshold, it is determined whether the target sample coating image meets the standard; In response to the target sample coating image conforming to the standard, the performance index parameters of the target sample within the historical period are collected; Analyze the performance index characterization values ​​based on the aforementioned performance index parameters; The difference between the performance index characterization value and the predetermined performance index characterization threshold is used to determine whether the coating quality of the target sample meets the standard. In response to the fact that the coating quality of the target sample does not meet the standard, the corresponding processing strategy is determined based on the difference between the performance index characterization value and the predetermined performance index characterization threshold. The image index parameters include the particle distribution density, coating thickness, and surface roughness. The performance parameters include porosity and coating tensile strength.

[0008] Furthermore, the process of analyzing the image index representation values ​​using the image index parameters includes: Collect data on the distribution density of sprayed particles, coating thickness, and surface roughness of the target sample over a historical period. The ratio of the sprayed particle distribution density to the predetermined coating particle distribution density threshold is calculated as the first image-defined characterization parameter; The ratio of the coating thickness to a predetermined coating thickness threshold is calculated as the second image-defined characterization parameter; The ratio of a predetermined surface roughness threshold to surface roughness is calculated as a third image-defined characterization parameter; The summation of the first image-defined characterization parameter, the second image-defined characterization parameter, and the third image-defined characterization parameter is determined as the image index characterization value.

[0009] Furthermore, the process of determining whether the target sample coating image conforms to the standard by comparing the image index characterization value with the predetermined image index characterization threshold includes: Extract the comparison results between the image index representation value and the predetermined image index representation threshold; If the image index characterization value is greater than the predetermined image index characterization threshold, then the target sample coating image is determined to meet the standard.

[0010] Furthermore, the process of determining whether the target sample coating image does not meet the standard by comparing the image index characterization value with the predetermined image index characterization threshold includes: Extract the comparison results between the image index representation value and the predetermined image index representation threshold; If the image index characterization value is less than or equal to the predetermined image index characterization threshold, then the target sample coating image is determined to be non-compliant with the standard.

[0011] Furthermore, the process of analyzing the performance index parameters and their representative values ​​includes: Extract the porosity and coating tensile strength of the target sample over a historical period; The ratio of the predetermined porosity threshold to the porosity is calculated as the first characteristic-defined characterization parameter; The ratio of the coating tensile strength to a predetermined coating tensile strength threshold is used as the second characteristic-defined characterization parameter. The sum of the first feature-limited characterization parameter and the second feature-limited characterization parameter is determined as the performance index characterization value.

[0012] Furthermore, the process of determining whether the coating quality of the target sample meets the standard by the difference between the performance index characterization value and the predetermined performance index characterization threshold includes: Calculate the difference between the performance index value and the predetermined performance index threshold; If the difference between the performance index value and the predetermined performance index threshold is greater than the predetermined difference threshold, then the coating quality of the target sample is determined to meet the standard.

[0013] Furthermore, the process of determining whether the coating quality of the target sample does not meet the standard based on the difference between the performance index characterization value and the predetermined performance index characterization threshold includes: Calculate the difference between the performance index value and the predetermined performance index threshold; If the difference between the performance index value and the predetermined performance index threshold is less than or equal to the predetermined difference threshold, then the coating quality of the target sample is determined to be non-compliant with the standard.

[0014] Furthermore, in response to the target sample coating quality not meeting the standard, the process of determining the corresponding processing strategy based on the difference between the performance index characterization value and the predetermined performance index characterization threshold includes: Calculate the difference between the performance index value and the predetermined performance index threshold; The adjustment range of the image index characterization threshold is determined based on the difference between the performance index characterization value and the predetermined performance index characterization threshold.

[0015] Furthermore, the process of ensuring that the coating image and quality of the target sample meet the standards includes: Extract the comparison results between the image index representation value and the predetermined image index representation threshold; Calculate the difference between the performance index value and the predetermined performance index threshold; If the image index characterization value is greater than the predetermined image index characterization threshold and the difference between the performance index characterization value and the predetermined performance index characterization threshold is greater than the predetermined difference threshold, then the target sample coating image and quality are determined to meet the standard.

[0016] Furthermore, the system for the machine vision-based thermal spray coating quality inspection method includes: Acquisition module: It is used to acquire image index parameters of the target sample over a historical period, including sprayed particle distribution density, coating thickness, and surface roughness; and to acquire performance index parameters of the target sample over a historical period, including porosity and coating tensile strength. Analysis module: It is connected to the acquisition module and is used to analyze image index characterization values ​​based on image index parameters; and to analyze performance index characterization values ​​based on performance index parameters. Evaluation module: It is connected to the analysis module and is used to determine whether the coating image of the target sample meets the standard; If the image index characterization value is greater than the predetermined image index characterization threshold, then the target sample coating image is determined to meet the standard. If the image index characterization value is less than or equal to the predetermined image index characterization threshold, the target sample coating image is determined to be non-compliant with the standard. Decision module: It is connected to the evaluation module and is used to determine whether the coating quality of the target sample meets the standard in response to the target sample coating image conforming to the standard; If the difference between the performance index value and the predetermined performance index threshold is greater than the predetermined difference threshold, then the coating quality of the target sample is determined to meet the standard. If the difference between the performance index characterization value and the predetermined performance index characterization threshold is less than or equal to the predetermined difference threshold, then the coating quality of the target sample is determined to be non-compliant with the standard. The control module is connected to the decision module and is used to determine the adjustment range of the image index characterization threshold. The adjustment range of the image index characterization threshold is determined based on the difference between the performance index characterization value and the predetermined performance index characterization threshold. The image index parameters include the particle distribution density, coating thickness, and surface roughness. The performance parameters include porosity and coating tensile strength.

[0017] Compared with existing technologies, the advantages of this invention lie in providing a machine vision-based thermal spray coating quality inspection system and method. By quantitatively analyzing the visual characteristics of sprayed particle distribution density, coating thickness, and surface roughness, it not only achieves automated and accurate judgment of coating appearance quality, but also constructs a nonlinear mapping relationship between visual features and key performance parameters such as porosity and coating tensile strength based on deep learning, thereby forming an early diagnosis and warning capability for the coating's intrinsic quality. The system of this invention adopts a hierarchical detection architecture of visual initial screening and performance verification, significantly improving the allocation efficiency of detection resources through multimodal data fusion. When abnormal performance indicators are identified, the system can trace the quantitative correlation between visual features and performance parameters, providing data support for closed-loop control of spraying process parameters, ultimately forming an advanced quality control system integrating real-time visual monitoring, quality prediction and evaluation, and intelligent process decision-making.

[0018] In particular, this invention constructs a key parameter system through two dimensions: process and quality performance. Image parameters include sprayed particle distribution density, coating thickness, and surface roughness to directly characterize the coating formation quality, while performance parameters include porosity and coating tensile strength to reflect coating performance. Standardization is performed by using the ratio of measured values ​​to predetermined thresholds, effectively eliminating dimensional differences between multi-source data. Then, a comprehensive characterization value is obtained by linearly superimposing each defined characterization parameter. The calculation method of this invention combines the comprehensiveness of multi-parameter fusion with the simplicity of engineering applications, providing a clear and reliable operational basis for multi-dimensional quantitative evaluation of coating quality.

[0019] In particular, this invention constructs a complete quality judgment and control system. It achieves rapid screening of coating appearance quality by directly comparing image index values ​​with set thresholds, while simultaneously employing a difference comparison method between performance index values ​​and standard thresholds to ensure strict compliance with final product performance. Based on this, the system innovatively establishes a dynamic image threshold adjustment mechanism based on performance testing results, forming a closed-loop optimization system from terminal quality feedback to production process control. Ultimately, through the coordinated judgment of both image quality and product performance indicators, the accuracy and reliability of quality evaluation are comprehensively guaranteed, significantly improving the adaptive control capability and overall quality control level of the testing system, and providing reliable technical support for continuous optimization of the production process.

[0020] In particular, this invention constructs a modular quality control system centered on machine vision, including an acquisition module, an analysis module, an evaluation module, a decision-making module, and a control module. The integrated vision acquisition module acquires multi-dimensional image features of sprayed particle distribution density, coating thickness, and surface roughness, and establishes a correlation model with porosity and coating tensile strength performance parameters. The system adopts a hierarchical processing architecture: the vision analysis module completes feature extraction and fusion calculation of multi-source image data; the evaluation module achieves real-time judgment of appearance quality based on preset thresholds; the decision-making module ensures quality reliability through dynamic comparison of performance prediction values ​​with thresholds; and the control module innovatively optimizes the visual judgment threshold based on performance feedback data, forming a complete closed-loop control system for visual inspection, quality prediction, and process optimization, realizing a full-process intelligent upgrade from image feature extraction to intelligent quality decision-making. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the steps of the machine vision-based thermal spray coating quality inspection method according to an embodiment of the present invention. Figure 2 This invention provides a logic diagram for determining whether a visual image of a target sample conforms to a standard. Figure 3 This invention provides a logic diagram for determining whether the coating quality of a target sample meets the standard. Figure 4 This is a structural block diagram of a machine vision-based thermal spray coating quality inspection system according to an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0023] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0024] Please see Figure 1 The diagram shows a flowchart of the steps in a machine vision-based thermal spray coating quality inspection method according to an embodiment of the present invention. The present invention provides a machine vision-based thermal spray coating quality inspection method, comprising: Step S1: Collect image index parameters of the target sample within the historical period; Step S2: Analyze the image index representation value based on the image index parameters; Step S3: Determine whether the target sample coating image meets the standard based on the comparison result between the image index characterization value and the predetermined image index characterization threshold; Step S4: In response to the target sample coating image conforming to the standard, collect the performance index parameters of the target sample within the historical period; analyze the performance index characterization value based on the performance index parameters; Step S5: Determine whether the coating quality of the target sample meets the standard based on the difference between the performance index characterization value and the predetermined performance index characterization threshold. Step S6: In response to the fact that the coating quality of the target sample does not meet the standard, a corresponding processing strategy is determined based on the difference between the performance index characterization value and the predetermined performance index characterization threshold.

[0025] In this embodiment, image index parameters are obtained through machine vision. A high-speed camera and a backlight source are used to capture images of the sprayed particle flow sequence. After threshold segmentation and feature analysis, the number and proportion of particles per unit area are calculated to obtain the particle distribution density. The particle distribution density determines the corrosion resistance, bonding strength, and intrinsic quality of the coating. A laser displacement sensor scans the surface of the target sample before and after spraying. The real-time thickness of the coating is obtained directly by calculating the height difference between the two contours. Sufficient coating thickness is the guarantee for its anti-corrosion and wear-resistant functions, and at the same time, it provides the necessary material allowance for subsequent machining. A high-resolution area array camera is used to acquire images of the coating surface under low-angle grazing light conditions. The surface roughness is indirectly characterized by analyzing the statistical characteristics of the image grayscale values. The standard deviation of the image grayscale values ​​is calculated as the surface roughness, which is used to control the material allowance for subsequent precision machining to ensure the final thickness and integrity of the functional coating.

[0026] In this embodiment, by quantitatively analyzing the visual characteristics of sprayed particle distribution density, coating thickness, and surface roughness, not only is the automated and accurate determination of coating appearance quality achieved, but also a nonlinear mapping relationship between visual features and key performance parameters such as porosity and coating tensile strength is constructed based on deep learning. This enables early diagnosis and warning of the coating's intrinsic quality. The system of this invention adopts a hierarchical detection architecture of visual screening and performance verification, significantly improving the efficiency of detection resource allocation through multimodal data fusion. When abnormal performance indicators are identified, the system can trace the quantitative correlation between visual features and performance parameters, providing data support for closed-loop control of spraying process parameters. Ultimately, this forms an advanced quality control system integrating real-time visual monitoring, quality prediction and evaluation, and intelligent process decision-making.

[0027] Specifically, the process of analyzing the image index representation values ​​using the image index parameters includes: Collect data on the distribution density of sprayed particles, coating thickness, and surface roughness of the target sample over a historical period. The ratio of the sprayed particle distribution density to the predetermined coating particle distribution density threshold is calculated as the first image-defined characterization parameter; The ratio of the coating thickness to a predetermined coating thickness threshold is calculated as the second image-defined characterization parameter; The ratio of a predetermined surface roughness threshold to surface roughness is calculated as a third image-defined characterization parameter; The summation of the first image-defined characterization parameter, the second image-defined characterization parameter, and the third image-defined characterization parameter is determined as the image index characterization value.

[0028] In this embodiment, the predetermined coating particle distribution density threshold, coating thickness threshold, and surface roughness threshold are all obtained in advance. The coating particle distribution density, coating thickness, and surface roughness of the target sample during thermal spraying are collected over a three-month period, and their average values ​​are calculated and determined as the predetermined coating particle distribution density threshold, coating thickness threshold, and surface roughness threshold.

[0029] In this embodiment, the formula for calculating surface roughness is:

[0030] Where R is the surface roughness; I(i,j) is the gray intensity value of pixel (i,j); M and N are the number of rows and columns of the image; This represents the average grayscale value of the entire image.

[0031] In this embodiment, three key image parameters—spray particle distribution density, coating thickness, and surface roughness—are collected. These parameters are then calculated as ratios to predetermined thresholds to obtain three standardized image-defined characterization parameters. Finally, a comprehensive image index characterization value is obtained by summing these parameters. The ratio calculation effectively eliminates dimensional differences between different parameters, enabling multi-source heterogeneous data to be fused and directly comparable. The three parameters characterize different dimensions of coating quality, and their weighted fusion can comprehensively reflect the overall appearance of the coating. The linear summation algorithm is simple and clear, facilitating engineering implementation and real-time calculation.

[0032] Please see Figure 2 As shown, this is a logic diagram for determining whether a target sample visual image conforms to a standard, provided by an embodiment of the present invention. The process for determining whether a target sample visual image conforms to a standard provided by an embodiment of the present invention includes: Extract the comparison results between the image index representation value and the predetermined image index representation threshold; If the image index characterization value is greater than the predetermined image index characterization threshold, then the target sample coating image is determined to meet the standard. If the image index characterization value is less than or equal to the predetermined image index characterization threshold, then the target sample coating image is determined to be non-compliant with the standard.

[0033] In this embodiment, the predetermined image index characterization threshold is obtained in advance. The image index characterization values ​​of the target sample during thermal spraying are collected over a three-month period, and their average value is calculated and determined as the predetermined image index characterization threshold. In this embodiment, the predetermined image index characterization threshold is selected within the range [3.05, 3.35]. Preferably, the predetermined image index characterization threshold is 3.15. By setting the predetermined image index characterization threshold and comparing and judging, an objective and unified anomaly judgment standard is established.

[0034] In this embodiment, a clear and explicit judgment criterion is established by directly comparing the image index characterization value obtained through comprehensive calculation with a pre-set characterization threshold. When the image index characterization value is greater than the predetermined threshold, it is judged to meet the standard; otherwise, it is judged to not meet the standard. This achieves complete automation and objectivity of the judgment process, effectively avoiding subjective bias in human judgment and ensuring the consistency and repeatability of the detection results. The threshold comparison algorithm based on quantitative data is simple to calculate and has high execution efficiency, which can meet the real-time detection needs of industrial production. The judgment criteria of this embodiment are clearly defined and highly operable, providing a clear execution basis for quality control on the production site and greatly improving the standardization and decision-making efficiency of quality inspection.

[0035] Specifically, the process of analyzing the performance index parameters and their representative values ​​includes: Extract the porosity and coating tensile strength of the target sample over a historical period; The ratio of the predetermined porosity threshold to the porosity is calculated as the first characteristic-defined characterization parameter; The ratio of the coating tensile strength to a predetermined coating tensile strength threshold is used as the second characteristic-defined characterization parameter. The sum of the first feature-limited characterization parameter and the second feature-limited characterization parameter is determined as the performance index characterization value.

[0036] In this embodiment, the predetermined porosity threshold and coating tensile strength threshold are obtained in advance. The porosity and coating tensile strength of the target sample after thermal spraying are collected over a three-month period, and their average values ​​are calculated and determined as the predetermined porosity threshold and coating tensile strength threshold.

[0037] In this embodiment, porosity can be determined using metallographic methods. Metallographic methods involve grinding and polishing the cross-sectional sample of the coating and then using digital image analysis technology to calculate the percentage of the area occupied by pores. The tensile bond strength of the coating is determined according to the ASTM C633 standard test method. A high-strength structural adhesive is used to bond the coating sample to a mating fixture, and an axial tensile load is applied on a universal testing machine until failure. The bond strength value is calculated by the ratio of the maximum failure load to the bonded area. Both methods provide internationally standardized quantitative data for coating performance and are indispensable benchmarks for establishing a model linking machine vision features and the intrinsic quality of the coating.

[0038] In this embodiment, the formula for calculating porosity is: Porosity = Pore area / Total area In this embodiment, the formula for calculating the tensile strength of the coating is: Tensile strength = maximum tensile force / cross-sectional area of ​​coating In this embodiment, porosity and coating tensile strength are selected as performance parameters to reflect coating performance. The ratio of the measured value to a predetermined threshold is used for standardization, which effectively eliminates the dimensional differences of multi-source data. Then, the comprehensive characterization value is obtained by linear superposition of each limited characterization parameter. The calculation method of this embodiment combines the comprehensiveness of multi-parameter fusion with the simplicity of engineering application, providing a clear and reliable operational basis for multi-dimensional quantitative evaluation of coating quality.

[0039] Please see Figure 3 As shown, this is a logic diagram for determining whether the coating quality of a target sample meets the standard, provided by an embodiment of the present invention. The process for determining whether the coating quality of a target sample meets the standard includes: Calculate the difference between the performance index value and the predetermined performance index threshold; If the difference between the performance index value and the predetermined performance index threshold is greater than the predetermined difference threshold, then the coating quality of the target sample is determined to meet the standard. If the difference between the performance index value and the predetermined performance index threshold is less than or equal to the predetermined difference threshold, then the coating quality of the target sample is determined to be non-compliant with the standard.

[0040] In this embodiment, the predetermined performance index characterization threshold is obtained in advance. The performance index characterization values ​​of the target sample after thermal spraying treatment are collected within a three-month period, and their average value is calculated. The predetermined performance index characterization threshold in this embodiment is selected within the range [2.05, 2.35]. Preferably, the predetermined performance index characterization threshold in this embodiment is 2.15.

[0041] In this embodiment, the predetermined difference threshold is obtained in advance. The difference between the performance index characterization value of the target sample after thermal spraying treatment and the predetermined performance index characterization threshold is calculated within a three-month period, and the average value is calculated as the predetermined difference threshold. In this embodiment, the predetermined difference threshold is selected within the range [0.05, 0.35]. Preferably, the predetermined difference threshold is 0.15.

[0042] In this embodiment, a more rigorous gradient quality assessment system is constructed by calculating the quantitative difference between the performance index characterization value and a predetermined threshold, and setting an independent difference threshold as a judgment benchmark. The judgment logic of this embodiment not only realizes the accurate quantitative assessment of the coating's intrinsic performance, but also effectively identifies the extent of performance index exceedance through the difference comparison mechanism. This effectively avoids the problem of misjudging the critical state that may exist in the traditional single threshold judgment, significantly improves the sensitivity and reliability of quality detection, provides accurate data support for the performance grading and quality traceability of coating products, and establishes a scientific decision-making basis for subsequent process parameter optimization.

[0043] Specifically, the process of determining the corresponding processing strategy based on the difference between the performance index characterization value and the predetermined performance index characterization threshold in response to the target sample coating quality not meeting the standard includes: Calculate the difference between the performance index value and the predetermined performance index threshold; The adjustment range of the image index characterization threshold is determined based on the difference between the performance index characterization value and the predetermined performance index characterization threshold.

[0044] In this embodiment, by establishing an adaptive feedback control mechanism, when the coating performance is detected to be substandard, the system will dynamically adjust the image index characterization threshold based on the quantitative difference between the performance index characterization value and the threshold, thereby realizing feedback closed-loop control from the terminal quality result to the process detection standard, enabling the system to have the ability to continuously self-optimize, and effectively upgrading quality control from passive detection to proactive prevention.

[0045] Specifically, the process of ensuring that the target sample coating image and quality meet the standards includes: Extract the comparison results between the image index representation value and the predetermined image index representation threshold; Calculate the difference between the performance index value and the predetermined performance index threshold; If the image index characterization value is greater than the predetermined image index characterization threshold and the difference between the performance index characterization value and the predetermined performance index characterization threshold is greater than the predetermined difference threshold, then the target sample coating image and quality are determined to meet the standard.

[0046] In this embodiment, the machine vision system achieves intelligent evaluation of coating quality by constructing a dual compliance judgment architecture of images and performance parameters. The judgment logic of the system in this embodiment requires that two core conditions be met simultaneously: the image index characterization value based on machine vision analysis is greater than a preset threshold, and the difference between the performance index characterization value calculated by the correlation model and the standard threshold meets the specified requirements. The collaborative judgment mechanism breaks through the limitation of traditional visual inspection focusing only on appearance features, enabling the machine vision system to not only complete the automated evaluation of appearance quality, but also effectively predict the intrinsic performance of the coating. This significantly improves the decision-making level of machine vision in industrial inspection and provides a complete solution for coating quality control that combines appearance inspection and performance prediction capabilities.

[0047] Please see Figure 4 The diagram shown is a structural block diagram of a machine vision-based thermal spray coating quality inspection system according to an embodiment of the present invention. The present invention also provides a machine vision-based thermal spray coating quality inspection system, comprising: Acquisition module: It is used to acquire image index parameters of the target sample over a historical period, including sprayed particle distribution density, coating thickness, and surface roughness; and to acquire performance index parameters of the target sample over a historical period, including porosity and coating tensile strength. Analysis module: It is connected to the acquisition module and is used to analyze image index characterization values ​​based on image index parameters; and to analyze performance index characterization values ​​based on performance index parameters. Evaluation module: It is connected to the analysis module and is used to determine whether the coating image of the target sample meets the standard; If the image index characterization value is greater than the predetermined image index characterization threshold, then the target sample coating image is determined to meet the standard. If the image index characterization value is less than or equal to the predetermined image index characterization threshold, the target sample coating image is determined to be non-compliant with the standard. Decision module: It is connected to the evaluation module and is used to determine whether the coating quality of the target sample meets the standard in response to the target sample coating image conforming to the standard; If the difference between the performance index value and the predetermined performance index threshold is greater than the predetermined difference threshold, then the coating quality of the target sample is determined to meet the standard. If the difference between the performance index characterization value and the predetermined performance index characterization threshold is less than or equal to the predetermined difference threshold, then the coating quality of the target sample is determined to be non-compliant with the standard. The control module is connected to the decision module and is used to determine the adjustment range of the image index characterization threshold. The adjustment range of the image index characterization threshold is determined based on the difference between the performance index characterization value and the predetermined performance index characterization threshold. The image index parameters include the particle distribution density, coating thickness, and surface roughness. The performance parameters include porosity and coating tensile strength.

[0048] In this embodiment, a modular quality control system centered on machine vision is constructed, including an acquisition module, an analysis module, an evaluation module, a decision-making module, and a control module. The integrated vision acquisition module acquires multi-dimensional image features of sprayed particle distribution density, coating thickness, and surface roughness, and establishes a correlation model with porosity and coating tensile strength performance parameters. The system adopts a hierarchical processing architecture. The vision analysis module completes feature extraction and fusion calculation of multi-source image data. The evaluation module realizes real-time judgment of appearance quality based on preset thresholds. The decision-making module ensures quality reliability by dynamically comparing performance prediction values ​​with thresholds. The control module innovatively optimizes the visual judgment threshold based on performance feedback data, forming a complete closed-loop control system for visual inspection, quality prediction, and process optimization, realizing a full-process intelligent upgrade from image feature extraction to intelligent quality decision-making.

[0049] The technical solution of the present invention has been described in conjunction with the embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to the specific implementation methods of the embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A machine vision-based method for inspecting the quality of thermal spray coatings, characterized in that, include: Collect image index parameters of the target sample within a historical period; Analyze the image index characterization values ​​based on the aforementioned image index parameters; Based on the comparison between the image index characterization value and the predetermined image index characterization threshold, it is determined whether the target sample coating image meets the standard; In response to the target sample coating image conforming to the standard, the performance index parameters of the target sample within the historical period are collected; Analyze the performance index characterization values ​​based on the aforementioned performance index parameters; The difference between the performance index characterization value and the predetermined performance index characterization threshold is used to determine whether the coating quality of the target sample meets the standard. In response to the fact that the coating quality of the target sample does not meet the standard, the corresponding processing strategy is determined based on the difference between the performance index characterization value and the predetermined performance index characterization threshold. The image index parameters include the particle distribution density, coating thickness, and surface roughness. The performance parameters include porosity and coating tensile strength.

2. The machine vision-based thermal spray coating quality inspection method according to claim 1, characterized in that, The process of analyzing image indicator representation values ​​based on the aforementioned image indicator parameters includes: Collect data on the distribution density of sprayed particles, coating thickness, and surface roughness of the target sample over a historical period. The ratio of the sprayed particle distribution density to the predetermined coating particle distribution density threshold is calculated as the first image-defined characterization parameter; The ratio of the coating thickness to a predetermined coating thickness threshold is calculated as the second image-defined characterization parameter; The ratio of a predetermined surface roughness threshold to surface roughness is calculated as a third image-defined characterization parameter; The summation of the first image-defined characterization parameter, the second image-defined characterization parameter, and the third image-defined characterization parameter is determined as the image index characterization value.

3. The machine vision-based thermal spray coating quality inspection method according to claim 2, characterized in that, The process of determining whether a target sample coating image conforms to a standard based on the comparison between the image index characterization value and a predetermined image index characterization threshold includes: Extract the comparison results between the image index representation value and the predetermined image index representation threshold; If the image index characterization value is greater than the predetermined image index characterization threshold, then the target sample coating image is determined to meet the standard.

4. The machine vision-based thermal spray coating quality inspection method according to claim 3, characterized in that, The process of determining whether a target sample coating image does not meet the standard based on the comparison result between the image index characterization value and the predetermined image index characterization threshold includes: Extract the comparison results between the image index representation value and the predetermined image index representation threshold; If the image index characterization value is less than or equal to the predetermined image index characterization threshold, then the target sample coating image is determined to be non-compliant with the standard.

5. The machine vision-based thermal spray coating quality inspection method according to claim 4, characterized in that, The process of analyzing the performance index characterization value based on the aforementioned performance index parameters includes: Extract the porosity and coating tensile strength of the target sample over a historical period; The ratio of the predetermined porosity threshold to the porosity is calculated as the first characteristic-defined characterization parameter; The ratio of the coating tensile strength to a predetermined coating tensile strength threshold is used as the second characteristic-defined characterization parameter. The sum of the first feature-limited characterization parameter and the second feature-limited characterization parameter is determined as the performance index characterization value.

6. The machine vision-based thermal spray coating quality inspection method according to claim 5, characterized in that, The process of determining whether the coating quality of the target sample meets the standard based on the difference between the performance index characterization value and the predetermined performance index characterization threshold includes: Calculate the difference between the performance index value and the predetermined performance index threshold; If the difference between the performance index value and the predetermined performance index threshold is greater than the predetermined difference threshold, then the coating quality of the target sample is determined to meet the standard.

7. The machine vision-based thermal spray coating quality inspection method according to claim 6, characterized in that, The process of determining whether the coating quality of the target sample does not meet the standard based on the difference between the performance index characterization value and the predetermined performance index characterization threshold includes: Calculate the difference between the performance index value and the predetermined performance index threshold; If the difference between the performance index value and the predetermined performance index threshold is less than or equal to the predetermined difference threshold, then the coating quality of the target sample is determined to be non-compliant with the standard.

8. The machine vision-based thermal spray coating quality inspection method according to claim 7, characterized in that, The process of determining the corresponding processing strategy based on the difference between the performance index characterization value and the predetermined performance index characterization threshold in response to the target sample coating quality not meeting the standard includes: Calculate the difference between the performance index value and the predetermined performance index threshold; The adjustment range of the image index characterization threshold is determined based on the difference between the performance index characterization value and the predetermined performance index characterization threshold.

9. The machine vision-based thermal spray coating quality inspection method according to claim 8, characterized in that, The process of ensuring that the coating image and quality of the target sample meet the standards includes: Extract the comparison results between the image index representation value and the predetermined image index representation threshold; Calculate the difference between the performance index value and the predetermined performance index threshold; If the image index characterization value is greater than the predetermined image index characterization threshold, and the difference between the performance index characterization value and the predetermined performance index characterization threshold is greater than the predetermined difference threshold, then the target sample coating image and quality are determined to meet the standard.

10. A system for inspecting the quality of thermal spray coatings based on machine vision as described in claims 1 to 9, characterized in that, include: Acquisition module: It is used to acquire image index parameters of the target sample over a historical period, including the distribution density of sprayed particles, coating thickness, and surface roughness; Collect performance parameters of the target sample over a historical period, including porosity and coating tensile strength; Analysis module: It is connected to the acquisition module and is used to analyze the image index characterization values ​​based on image index parameters; Analyze the performance index characterization values ​​based on performance index parameters; Evaluation module: It is connected to the analysis module and is used to determine whether the coating image of the target sample meets the standard; If the image index characterization value is greater than the predetermined image index characterization threshold, then the target sample coating image is determined to meet the standard. If the image index characterization value is less than or equal to the predetermined image index characterization threshold, the target sample coating image is determined to be non-compliant with the standard. Decision module: It is connected to the evaluation module and is used to determine whether the coating quality of the target sample meets the standard in response to the target sample coating image conforming to the standard; If the difference between the performance index value and the predetermined performance index threshold is greater than the predetermined difference threshold, then the coating quality of the target sample is determined to meet the standard. If the difference between the performance index characterization value and the predetermined performance index characterization threshold is less than or equal to the predetermined difference threshold, then the coating quality of the target sample is determined to be non-compliant with the standard. The control module is connected to the decision module and is used to determine the adjustment range of the image index characterization threshold. The adjustment range of the image index characterization threshold is determined based on the difference between the performance index characterization value and the predetermined performance index characterization threshold. The image index parameters include the particle distribution density, coating thickness, and surface roughness. The performance parameters include porosity and coating tensile strength.

Images