Plate defect detection method

Abstract

The invention belongs to the technical field of plate detection, and particularly relates to a plate defect detection method which comprises the steps of plate conveying, plate cleaning, plate image acquisition, image processing and defect identification and plate grade division. A plate to be detected is conveyed through the stable conveying module, when the plate enters a working area of the plate cleaning module, the surface of the plate is cleaned, and after cleaning is completed, a surface image of the plate is collected through the high-definition image collecting module; the image processing identification module carries out preprocessing and defect identification classification on the surface image of the plate, and the plate quality evaluation module carries out comprehensive evaluation on the quality condition of the plate based on a defect identification classification result to determine the grade of the plate, so that efficient and accurate detection on the surface defect of the plate is realized, and the purpose of improving the quality control and production efficiency of plate production is achieved. And the workload and the operation difficulty of operators are obviously reduced.

Description

Technical Field

[0001] The invention relates to the technical field of plate detection, and in particular to a plate defect detection method. Background Art

[0002] As an important industrial raw material, sheet metal is widely used in many fields such as construction, automobile, aerospace, furniture manufacturing, etc. However, sheet metal is prone to various surface defects during production and processing. These defects not only affect the appearance of the product, but also reduce its performance and safety.

[0003] A Chinese invention patent with publication number CN113362276A discloses a visual inspection method and system for plates. The technical solution of the invention is to use a 3D laser acquisition technology and a 3D profiler to obtain 3D point cloud information of the plate, integrate and display the collected 3D point cloud information to form a 3D model, and analyze the 3D model to identify the size defects of the target plate, eliminating the most primitive manual inspection method.

[0004] However, in actual application, the above-mentioned technical solution fails to clean the plate surface before image acquisition and processing and effectively monitor and automatically and adaptively regulate the cleaning process, and cannot accurately evaluate the degree of cleaning hazards and image acquisition hazards and issue timely warnings, which is not conducive to ensuring the high efficiency and stability of the plate detection process and the accuracy of the detection results;

[0005] In view of the above technical defects, a solution is now proposed. Summary of the invention

[0006] The object of the present invention is to provide a plate defect detection method to solve the technical defects proposed by the background technology.

[0007] To achieve the above object, the present invention provides the following technical solution: a plate defect detection method, comprising the following steps:

[0008] Step 1: The operation control module controls the stable transmission module, and the stable transmission module transports the plate to be inspected;

[0009] Step 2: The operation control module controls the plate cleaning module, and the plate cleaning module cleans the surface of the plate by rotating brush sweeping and blowing air;

[0010] Step 3: After the cleaning is completed, the plate enters the lower part of the high-definition imaging module. The high-definition imaging module collects the surface image of the plate through the industrial camera and transmits the surface image of the plate to the image processing and recognition module in real time;

[0011] Step 4: The image processing and recognition module pre-processes the surface image of the plate, extracts the defect features of the plate using edge detection and texture analysis algorithms, classifies the features and marks the defect types based on the deep learning model, and sends the plate defect recognition information to the plate material evaluation module;

[0012] Step 5: The board material evaluation module comprehensively evaluates the quality of the board, classifies the board into grades, and sends the grade classification information of the board to the display terminal.

[0013] Furthermore, the specific process of the plate material evaluation module to grade the plates is as follows:

[0014] All defects existing in the plate are obtained, and the detection data of various defect parameters of the corresponding defects are collected. The detection data of various defect parameters are compared with the corresponding preset data thresholds. If there are defect parameters whose detection data exceeds the corresponding preset data thresholds, the corresponding defects are marked as dangerous defects; if there are dangerous defects in the plate, the corresponding plate is marked as a low-grade plate;

[0015] If there are no dangerous defects in the plate, the distribution areas of all defects are summed up to obtain the defect distribution value, and the maximum deviation amplitude of the thickness, the maximum deviation amplitude of the width and the maximum deviation amplitude of the length of the plate are collected and marked as the thickness deviation detection value, the width deviation detection value and the length deviation detection value respectively;

[0016] The quality assessment coefficient is obtained by numerically calculating the defect distribution value, thickness deviation detection value, width deviation detection value and length deviation detection value, and the quality assessment coefficient is numerically compared with the preset quality assessment coefficient threshold. If the quality assessment coefficient exceeds the preset quality assessment coefficient threshold, the corresponding plate is marked as a low-grade plate; if the quality assessment coefficient does not exceed the preset quality assessment coefficient threshold, the corresponding plate is marked as a high-grade plate.

[0017] Furthermore, the operation control module is communicatively connected to the cleaning monitoring module. When cleaning the plate, the cleaning monitoring module monitors the cleaning process of the plate and analyzes abnormal performance of the cleaning process to determine whether to generate a cleaning control signal and send the cleaning control signal to the operation control module. When the operation control module receives the cleaning control signal, it adaptively controls the smooth conveying module and the plate cleaning module.

[0018] Furthermore, the specific analysis process of cleaning the monitoring module includes:

[0019] The rotation speed of the rotating brush roller in the plate cleaning module is collected and its deviation value compared with the set standard rotation speed is marked as the brush rotation speed value, and the pressure applied by the rotating brush roller to the plate during brushing is collected and its deviation value compared with the set standard pressure is marked as the brush extrusion value;

[0020] The blowing speed and blowing temperature of the air blowing device in the plate cleaning module are collected, and the deviation value of the blowing speed compared to the set standard blowing speed is marked as the blowing speed value. Similarly, the blowing temperature value is obtained, and the degree of deviation of the blowing direction of the air blowing device compared to the set standard direction is marked as the blowing direction value;

[0021] The brushing rotation speed value, brushing extrusion value, blowing speed value, blowing temperature value and blowing direction value are numerically compared with the preset brushing rotation speed threshold, the preset brushing extrusion threshold, the preset blowing speed threshold, the preset blowing temperature threshold and the preset blowing direction threshold respectively. If the brushing rotation speed value, the brushing extrusion value, the blowing speed value, the blowing temperature value or the blowing direction value exceeds the corresponding preset threshold, a cleaning control signal is generated.

[0022] Further, if the brushing rotation speed value, the brushing extrusion value, the blasting speed value, the blasting temperature value and the blasting direction value do not exceed the corresponding preset threshold value, the brushing rotation speed value, the brushing extrusion value, the blasting speed value, the blasting temperature value and the blasting direction value are respectively assigned corresponding preset weight coefficients, and the brushing rotation speed value, the brushing extrusion value, the blasting speed value, the blasting temperature value and the blasting direction value are respectively multiplied by the corresponding preset weight coefficients, and the sum of the five groups of multiplication results is marked as the cleaning analysis value;

[0023] And obtain the transport impact value through transport impact analysis, set several groups of preset transport impact value ranges in advance, and each group of preset transport impact value ranges corresponds to a group of preset impact factors; compare the transport impact value with all preset transport impact value ranges one by one, mark the preset transport impact value range containing the corresponding transport impact value as a matching range, and mark the impact factor corresponding to the matching range as a target factor; obtain the cleaning monitoring value by multiplying the cleaning analysis value with the target factor, compare the cleaning monitoring value with the preset cleaning monitoring threshold, and generate a cleaning control signal if the cleaning monitoring value exceeds the preset cleaning monitoring threshold.

[0024] Furthermore, the specific analysis process of the transport impact analysis is as follows:

[0025] The conveying speed of the smooth conveying module is collected and its deviation value compared with the set standard conveying speed is marked as the conveying speed detection value, and the shaking amplitude of the plate during the conveying process is collected and marked as the conveying shaking detection value. The conveying impact value is obtained by numerically calculating the conveying speed detection value and the conveying shaking detection value.

[0026] Furthermore, the cleaning monitoring module is communicatively connected to the cleaning hidden danger analysis module, and the cleaning monitoring module sends the cleaning control signal to the cleaning hidden danger analysis module. The cleaning hidden danger analysis module is used to set a detection period, and analyze the degree of cleaning hidden dangers for plate cleaning within the detection period. A high cleaning hidden danger signal or a low cleaning hidden danger signal is generated through analysis, and the high cleaning hidden danger signal is sent to the display terminal. The display terminal displays the high cleaning hidden danger signal and issues a corresponding warning.

[0027] Furthermore, the specific analysis process of cleaning hidden danger analysis module is as follows:

[0028] The number of times the cleaning control signal is generated during the detection period is collected and marked as the cleaning detection value, and the generation time of the cleaning control signal is marked as the cleaning time, and the time when the operation control module completes the corresponding control operation is marked as the adjustment completion time, and the interval between the adjustment completion time and the cleaning time is marked as the adjustment detection value;

[0029] Compare the inspection value with the preset inspection threshold. If the inspection value exceeds the preset inspection threshold, mark the corresponding inspection value as the risk adjustment value. Obtain the number of risk adjustment values ​​in the detection period and mark them as risk adjustment detection values. Compare the clearance detection value and the risk adjustment detection value with the preset clearance detection threshold and the preset risk adjustment detection threshold, respectively. If the clearance detection value or the risk adjustment detection value exceeds the corresponding preset threshold, generate a high risk clearance signal.

[0030] If both the cleaning and adjustment detection value and the risk adjustment detection value do not exceed the corresponding preset thresholds, then the average of all the risk adjustment values ​​in the detection period will be calculated to obtain the meter adjustment value, and the cleaning hidden danger coefficient will be obtained by numerically calculating the cleaning and adjustment detection value, the risk adjustment detection value and the meter adjustment value. The cleaning hidden danger coefficient is numerically compared with the preset cleaning hidden danger coefficient threshold. If the cleaning hidden danger coefficient exceeds the preset cleaning hidden danger coefficient threshold, a high hidden danger cleaning signal is generated; if the cleaning hidden danger coefficient does not exceed the preset cleaning hidden danger coefficient threshold, a low hidden danger cleaning signal is generated.

[0031] Furthermore, the hidden danger clearing analysis module is communicatively connected to the hidden danger collection analysis module, and the hidden danger clearing analysis module sends a low hidden danger clearing signal to the hidden danger collection analysis module. When the hidden danger collection analysis module receives the low hidden danger clearing signal, it analyzes the image collection performance of the high-definition imaging module within the detection period, and generates a high hidden danger collection signal or a low hidden danger collection signal through analysis, and sends the high hidden danger collection signal to the display terminal. When the display terminal receives the high hidden danger collection signal, it displays it and issues a corresponding warning.

[0032] Furthermore, the specific analysis process of the hidden danger analysis module is as follows:

[0033] The image resolution, color distortion and image noise value of the surface image captured by the high-definition imaging module are obtained. If the image resolution exceeds a preset image resolution threshold and the color distortion and image noise values ​​do not exceed a preset color distortion threshold and a preset image noise threshold, the corresponding surface image is marked as a qualified image. In other cases, the corresponding surface image is marked as an abnormal image.

[0034] The number of abnormal images collected by the high-definition imaging module during the detection period is obtained and the ratio is calculated with the total number of images collected to obtain the abnormal image share value, and the abnormal image share value is numerically compared with the preset abnormal image share threshold value. If the abnormal image share value exceeds the preset abnormal image share threshold value, a high hidden danger signal is generated; if the abnormal image share value does not exceed the preset abnormal image share threshold value, a low hidden danger signal is generated.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] 1. In the present invention, the plate to be inspected is transported by a stable conveying module, the plate cleaning module cleans the surface of the plate, and after the cleaning is completed, the surface image of the plate is collected by a high-definition imaging module, the image processing and recognition module pre-processes the surface image of the plate and classifies the defects, and the plate material evaluation module comprehensively evaluates the quality status of the plate based on the defect recognition and classification results to determine its grade, thereby realizing efficient and accurate detection of surface defects of the plate, and significantly reducing the workload and operation difficulty of operators;

[0037] 2. In the present invention, the cleaning process of the plate is effectively monitored and automatically and reasonably regulated through the cleaning monitoring module to ensure the cleaning effect of the plate and the smooth and stable progress of the cleaning process. The cleaning hidden danger analysis module analyzes the degree of cleaning hidden dangers within the detection period. When generating a cleaning low hidden danger signal, the hidden danger analysis module analyzes the image acquisition performance within the detection period, which is beneficial for the operator to conduct cause investigation and analysis in time and make reasonable improvement measures, further ensuring the subsequent detection efficiency and the stability of the detection process. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0039] Figure 1 This is a flow chart of the method of Embodiment 1 of the present invention;

[0040] Figure 2 This is a system block diagram of Embodiment 1 of the present invention;

[0041] Figure 3 This is a system block diagram of Embodiment 2, Embodiment 3 and Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] Embodiment 1: Figure 1-2 As shown, a plate defect detection method proposed by the present invention comprises the following steps:

[0044] Step 1: The operation control module controls the smooth conveying module, and the smooth conveying module conveys the plate to be inspected; preferably, the smooth conveying module conveys the plate by conveyor belt.

[0045] Step 2. The plate enters the working area of ​​the plate cleaning module, and the operation control module controls the plate cleaning module. The plate cleaning module (including a rotating brush roller and an air blowing device) cleans the surface of the plate by rotating brushes and air blowing to remove dust and other dirt on the surface of the plate, which is beneficial to the quality of the collected images and the accuracy of the defect recognition and analysis results.

[0046] Step 3: After cleaning, the plate enters the bottom of the high-definition imaging module. The high-definition imaging module collects the surface image of the plate through an industrial camera and transmits the surface image of the plate to the image processing and recognition module in real time.

[0047] Step 4: The image processing and recognition module preprocesses the surface image of the plate. The preprocessing includes denoising and contrast enhancement of the collected image to improve the accuracy of subsequent detection. The defect features of the plate are extracted using edge detection and texture analysis algorithms. The features are classified and the defect types (such as cracks, scratches, rust, color difference, pitting, peeling, oxidation, etc.) are marked based on deep learning models (such as convolutional neural networks (CNN)). The plate defect recognition information is sent to the plate material evaluation module.

[0048] Step 5: The plate material evaluation module comprehensively evaluates the quality status of the plate, grades the plate, and sends the grade classification information of the plate to the display terminal, which displays the grade classification information of the plate and the identified defect information (preferably, the operator can also manually control the operation control module and the image acquisition module through the display terminal, including parameter setting and modification, etc.), which is conducive to the operator to reasonably dispose of the plate, and can automatically detect and classify the surface defects of the plate and determine the grade of the plate. It has the advantages of fast detection speed and high accuracy, and realizes efficient and accurate detection of surface defects of the plate, aiming to improve the quality control and production efficiency of plate production; the process of plate material evaluation module to grade the plate is as follows:

[0049] All defects existing in the plate are obtained, and the detection data of various defect parameters of the corresponding defects (such as crack length, crack width, depth and other parameters) are collected. The detection data of various defect parameters are compared with the corresponding preset data thresholds. If there are defect parameters whose detection data exceeds the corresponding preset data thresholds (such as the crack width exceeds the preset crack width threshold, etc.), the corresponding defects are marked as dangerous defects; if there are dangerous defects in the plate, it indicates that the quality of the plate is poor, and the corresponding plate is marked as a low-grade plate;

[0050] If there are no dangerous defects in the plate, the distribution areas of all defects are summed up to obtain the defect distribution value, and the maximum deviation amplitude of the thickness, the maximum deviation amplitude of the width and the maximum deviation amplitude of the length of the plate are collected and marked as the thickness deviation detection value, the width deviation detection value and the length deviation detection value respectively;

[0051] By formula The defect distribution value WR, the thickness deviation detection value HP, the width deviation detection value CF and the length deviation detection value MG are numerically calculated to obtain the quality evaluation coefficient Y, wherein eq, bg, up, ng are preset weight coefficients with values ​​greater than zero, and the larger the value of the quality evaluation coefficient Y is, the worse the overall quality of the plate is;

[0052] The quality assessment coefficient Y is numerically compared with the preset quality assessment coefficient threshold. If the quality assessment coefficient Y exceeds the preset quality assessment coefficient threshold, it indicates that the quality condition of the plate is generally poor, and the corresponding plate is marked as a low-grade plate; if the quality assessment coefficient Y does not exceed the preset quality assessment coefficient threshold, it indicates that the quality condition of the plate is generally good, and the corresponding plate is marked as a high-grade plate.

[0053] Embodiment 2: Figure 3As shown, the difference between this embodiment and the first embodiment is that the operation control module is communicatively connected to the cleaning monitoring module. When the plate is cleaned, the cleaning monitoring module monitors the cleaning process of the plate, analyzes the abnormal performance of the cleaning process, and judges whether to generate a cleaning control signal based on this, and sends the cleaning control signal to the operation control module. When the operation control module receives the cleaning control signal, it adaptively controls the smooth conveying module and the plate cleaning module to achieve effective monitoring and automatic and reasonable control of the cleaning operation, ensure the cleaning effect for the plate, reduce the difficulty of operation and management of the operator, and facilitate the smooth and stable progress of the cleaning process; the specific analysis process of the cleaning monitoring module is as follows:

[0054] The rotation speed of the rotating brush roller in the plate cleaning module is collected and its deviation value compared with the set standard rotation speed is marked as the brush rotation speed value, and the pressure applied by the rotating brush roller to the plate during brushing is collected and its deviation value compared with the set standard pressure is marked as the brush extrusion value;

[0055] and collecting the blowing speed and blowing temperature of the air blowing device in the plate cleaning module, marking the deviation value of the blowing speed compared to the set standard blowing speed as the blowing speed value, marking the deviation value of the blowing temperature compared to the set standard blowing temperature as the blowing temperature value, and marking the degree of deviation of the blowing direction of the air blowing device compared to the set standard direction as the blowing direction value;

[0056] The brushing speed value, brushing extrusion value, blast speed value, blast temperature value and blast direction value are numerically compared with the preset brushing speed threshold, the preset brushing extrusion threshold, the preset blast speed threshold, the preset blast temperature threshold and the preset blast direction threshold. If the brushing speed value, the brushing extrusion value, the blast speed value, the blast temperature value or the blast direction value exceeds the corresponding preset threshold, it indicates that the cleaning operation condition of the plate is poor and corresponding adjustment is needed in time, then a cleaning adjustment signal is generated.

[0057] Further, if the brushing rotation speed value, the brushing extrusion value, the blasting speed value, the blasting temperature value and the blasting direction value do not exceed the corresponding preset threshold value, the brushing rotation speed value, the brushing extrusion value, the blasting speed value, the blasting temperature value and the blasting direction value are respectively assigned corresponding preset weight coefficients, and the brushing rotation speed value, the brushing extrusion value, the blasting speed value, the blasting temperature value and the blasting direction value are respectively multiplied by the corresponding preset weight coefficients, and the sum of the five groups of multiplication results is marked as the cleaning analysis value;

[0058] The conveying speed of the smooth conveying module is collected and its deviation from the set standard conveying speed is marked as the conveying speed detection value, and the shaking amplitude of the plate during the conveying process is collected and marked as the conveying shaking detection value. The conveying speed detection value GX and the conveying shaking detection value FY are numerically calculated to obtain the conveying influence value T; wherein sw and rw are preset weight coefficients with values ​​greater than zero, and the larger the value of the conveying influence value T is, the worse the real-time conveying condition of the plate by the stable conveying module is, and the more unfavorable it is for the effective cleaning of the plate;

[0059] A number of preset transport impact value ranges are set in advance, and each set of preset transport impact value ranges corresponds to a set of preset impact factors with values ​​greater than zero; it should be noted that the larger the value of the preset transport impact value range is, the larger the value of the preset impact factor that matches it is; the transport impact value is compared with all the preset transport impact value ranges one by one, the preset transport impact value range containing the corresponding transport impact value is marked as a matching range, and the impact factor corresponding to the matching range is marked as a target factor;

[0060] The cleaning analysis value T is multiplied by the target factor to obtain the cleaning monitoring value, and the cleaning monitoring value is numerically compared with the preset cleaning monitoring threshold. If the cleaning monitoring value exceeds the preset cleaning monitoring threshold, it indicates that the cleaning operation condition of the plate is generally poor and needs to be adjusted accordingly, then a cleaning control signal is generated.

[0061] Embodiment 3: Figure 3 As shown, the difference between this embodiment and the first and second embodiments is that the cleaning monitoring module is communicatively connected to the cleaning hidden danger analysis module, the cleaning monitoring module sends the cleaning control signal to the cleaning hidden danger analysis module, the cleaning hidden danger analysis module counts the generation frequency of the cleaning control signal, and the cleaning hidden danger analysis module is used to set the detection period, analyze the cleaning hidden danger degree of the plate cleaning within the detection period, and generate a cleaning high hidden danger signal or a cleaning low hidden danger signal through analysis;

[0062] The cleaning high hidden danger signal is sent to the display terminal, which displays the cleaning high hidden danger signal and issues a corresponding warning. When the operator receives the corresponding warning, he / she conducts cause investigation and analysis and makes reasonable improvement measures, which is conducive to the stable and smooth progress of the subsequent plate cleaning operation. The specific analysis process of the cleaning hidden danger analysis module is as follows:

[0063] The number of times the cleaning control signal is generated within the detection period is collected and marked as the cleaning control detection value, and the generation time of the cleaning control signal is marked as the cleaning time, and the time when the operation control module completes the corresponding control operation is marked as the adjustment completion time, and the interval between the adjustment completion time and the cleaning time is marked as the adjustment detection value; and the larger the value of the adjustment detection value is, the worse the adjustment efficiency of the operation control module for the corresponding control operation is;

[0064] The adjustment inspection value is numerically compared with the preset adjustment inspection threshold. If the adjustment inspection value exceeds the preset adjustment inspection threshold, it indicates that the adjustment efficiency of the operation control module for the corresponding adjustment operation is poor, and the corresponding adjustment inspection value is marked as the risk adjustment value; the number of risk adjustment values ​​in the detection period is obtained and marked as the risk adjustment detection value, and the cleaning adjustment detection value and the risk adjustment detection value are numerically compared with the preset cleaning adjustment detection threshold and the preset risk adjustment detection threshold respectively. If the cleaning adjustment detection value or the risk adjustment detection value exceeds the corresponding preset threshold, it indicates that the cleaning process for the plate cleaning in the detection period is not smooth and it is difficult to ensure the cleaning effect for the plate, and a cleaning high risk signal is generated;

[0065] If the clearance detection value and the risk adjustment detection value do not exceed the corresponding preset threshold, the average of all risk adjustment values ​​in the detection period is calculated to obtain the meter adjustment value, and the formula is used The cleaning and adjustment detection value QF, the risk adjustment detection value YF and the meter adjustment value SL are numerically calculated to obtain the cleaning hidden danger coefficient B; wherein, ry, kp, es are preset weight coefficients with values ​​greater than zero, and the larger the value of the cleaning hidden danger coefficient B, the more unsmooth the cleaning process for the plate cleaning during the detection period;

[0066] The cleaning hidden danger coefficient B is numerically compared with the preset cleaning hidden danger coefficient threshold. If the cleaning hidden danger coefficient B exceeds the preset cleaning hidden danger coefficient threshold, it indicates that the cleaning process for the plate cleaning during the detection period is not smooth, and it is difficult to ensure the cleaning effect of the plate, then a high cleaning hidden danger signal is generated; if the cleaning hidden danger coefficient B does not exceed the preset cleaning hidden danger coefficient threshold, it indicates that the cleaning process for the plate cleaning during the detection period is relatively smooth, then a low cleaning hidden danger signal is generated.

[0067] Embodiment 4: Figure 3 As shown, the difference between this embodiment and the first, second and third embodiments is that the hidden danger analysis module is communicatively connected to the hidden danger collection analysis module, the hidden danger analysis module sends the low hidden danger cleaning signal to the hidden danger collection analysis module, and the hidden danger collection analysis module analyzes the image collection performance of the high-definition imaging module within the detection period when receiving the low hidden danger cleaning signal, and generates a high hidden danger collection signal or a low hidden danger collection signal through analysis;

[0068] The high hidden danger signal is sent to the display terminal. When the display terminal receives the high hidden danger signal, it will be displayed and a corresponding warning will be issued. When the operator receives the corresponding warning, he will conduct a cause investigation and analysis and make reasonable improvement measures for the image acquisition module, which is conducive to ensuring the quality of the subsequent collected images and improving the subsequent detection efficiency and image acquisition performance. The specific analysis process of the hidden danger analysis module is as follows:

[0069] The image resolution, color distortion and image noise value of the surface image captured by the high-definition imaging module are obtained, and the image resolution, color distortion and image noise values ​​are numerically compared with a preset image resolution threshold, a preset color distortion threshold and a preset image noise threshold, respectively; if the image resolution exceeds the preset image resolution threshold and the color distortion and image noise values ​​do not exceed the preset color distortion threshold and the preset image noise threshold, the corresponding surface image is marked as a qualified image; otherwise, the corresponding surface image is marked as an abnormal image;

[0070] The number of abnormal images collected by the high-definition imaging module during the detection period is obtained and the ratio is calculated with the total number of images collected to obtain the abnormal image share value, and the abnormal image share value is numerically compared with the preset abnormal image share threshold value. If the abnormal image share value exceeds the preset abnormal image share threshold value, it indicates that the image acquisition performance of the high-definition imaging module during the detection period is poor, and a high-risk acquisition signal is generated; if the abnormal image share value does not exceed the preset abnormal image share threshold value, it indicates that the image acquisition performance of the high-definition imaging module during the detection period is good, and a low-risk acquisition signal is generated.

[0071] When the present invention is in use, the plate to be inspected is transported by a smooth conveying module. When the plate enters the working area of ​​the plate cleaning module, the plate cleaning module cleans the surface of the plate by a combination of rotating brushing and air blowing to remove dust and other dirt on the surface of the plate, which is beneficial to ensure the quality of the collected image and the accuracy of the defect recognition and analysis results. After the cleaning is completed, the surface image of the plate is collected by the high-definition imaging module, and the image processing and recognition module pre-processes and classifies the surface image of the plate. The plate material evaluation module comprehensively evaluates the quality status of the plate based on the defect recognition and classification results to determine its grade, and can automatically detect and classify the surface defects of the plate and determine the grade of the plate, thereby realizing efficient and accurate detection of surface defects of the plate, aiming to improve the quality control and production efficiency of plate production, and significantly reduce the workload and operation difficulty of operators.

[0072] The above formulas are all dimensionless and numerical calculations. The formula is a formula obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formula are set by technicians in this field according to actual conditions. The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the contents of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that technicians in the relevant technical field can understand and use the present invention well. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A plate defect detection method, characterized in that: The following steps are involved: Step 1: The operation control module controls the stable transmission module, and the stable transmission module transports the plate to be inspected; Step 2: The operation control module controls the plate cleaning module, and the plate cleaning module cleans the surface of the plate by rotating brush sweeping and blowing air; Step 3: After the cleaning is completed, the plate enters the lower part of the high-definition imaging module. The high-definition imaging module collects the surface image of the plate through the industrial camera and transmits the surface image of the plate to the image processing and recognition module in real time; Step 4: The image processing and recognition module pre-processes the surface image of the plate, extracts the defect features of the plate using edge detection and texture analysis algorithms, classifies the features and marks the defect types based on the deep learning model, and sends the plate defect recognition information to the plate material evaluation module; Step 5: The board material evaluation module comprehensively evaluates the quality of the board, classifies the board into grades, and sends the grade classification information of the board to the display terminal.

2. A plate defect detection method according to claim 1, characterized in that: In step 5, the specific process of the plate material evaluation module to grade the plate is as follows: All defects existing in the plate are obtained. If there are dangerous defects in the plate, the corresponding plate is marked as a low-grade plate; if there are no dangerous defects in the plate, the quality evaluation coefficient is obtained by numerically calculating the defect distribution value, thickness deviation detection value, width deviation detection value and length deviation detection value. If the quality evaluation coefficient exceeds the preset quality evaluation coefficient threshold, the corresponding plate is marked as a low-grade plate; otherwise, the corresponding plate is marked as a high-grade plate.

3. A plate defect detection method according to claim 1, characterized in that: The operation control module is communicatively connected to the cleaning monitoring module. When cleaning the plate, the cleaning monitoring module monitors the cleaning process of the plate and analyzes the abnormal performance of the cleaning process, and determines whether to generate a cleaning control signal based on this, and sends the cleaning control signal to the operation control module. When the operation control module receives the cleaning control signal, it adaptively controls the smooth conveying module and the plate cleaning module.

4. A plate defect detection method according to claim 3, characterized in that: The specific analysis process of cleaning the monitoring module includes: If the brushing rotation speed value, the brushing extrusion value, the blowing speed value, the blowing temperature value or the blowing direction value exceeds the corresponding preset threshold value, a cleaning control signal is generated.

5. A plate defect detection method according to claim 4, characterized in that: If the brushing rotation speed value, brushing extrusion value, blast speed value, blast temperature value and blast direction value do not exceed the corresponding preset threshold values, the transport impact value is obtained through transport impact analysis, and the preset transport impact value range containing the corresponding transport impact value is marked as the matching range, and the impact factor corresponding to the matching range is marked as the target factor; the cleaning analysis value is multiplied by the target factor to obtain the cleaning monitoring value, and if the cleaning monitoring value exceeds the preset cleaning monitoring threshold, a cleaning control signal is generated.

6. A plate defect detection method according to claim 5, characterized in that: The specific analysis process of transport impact analysis is as follows: The conveying speed of the smooth conveying module is collected and its deviation value compared with the set standard conveying speed is marked as the conveying speed detection value, and the shaking amplitude of the plate during the conveying process is collected and marked as the conveying shaking detection value. The conveying impact value is obtained by numerically calculating the conveying speed detection value and the conveying shaking detection value.

7. A plate defect detection method according to claim 3, characterized in that: The cleaning monitoring module is communicatively connected to the cleaning hidden danger analysis module. The cleaning monitoring module sends the cleaning control signal to the cleaning hidden danger analysis module. The cleaning hidden danger analysis module is used to set the detection time period, analyze the degree of cleaning hidden dangers for plate cleaning within the detection period, and generate a high cleaning hidden danger signal or a low cleaning hidden danger signal through analysis, and send the high cleaning hidden danger signal to the display terminal.

8. A plate defect detection method according to claim 7, characterized in that: The specific analysis process of cleaning hidden danger analysis module is as follows: If the clearing and adjustment detection value or the risk adjustment detection value exceeds the corresponding preset threshold, a high hidden danger clearing signal is generated; if both the clearing and adjustment detection value and the risk adjustment detection value do not exceed the corresponding preset threshold, the hidden danger clearing coefficient is obtained by numerically calculating the clearing and adjustment detection value, the risk adjustment detection value and the meter adjustment value. If the hidden danger clearing coefficient exceeds the preset hidden danger clearing coefficient threshold, a high hidden danger clearing signal is generated; otherwise, a low hidden danger clearing signal is generated.

9. A plate defect detection method according to claim 7, characterized in that: The hidden danger clearing analysis module is communicatively connected to the hidden danger collection analysis module. The hidden danger clearing analysis module sends a low hidden danger clearing signal to the hidden danger collection analysis module. When the hidden danger collection analysis module receives the low hidden danger clearing signal, it analyzes the image collection performance of the high-definition imaging module during the detection period; if the abnormal image occupancy value exceeds the preset abnormal image occupancy threshold, a high hidden danger collection signal is generated, otherwise a low hidden danger collection signal is generated, and the high hidden danger collection signal is sent to the display terminal.

Images