A stamping sheet metal defect detection system and method based on infrared imaging

Abstract

The present invention relates to the technical field of plate defect detection, and discloses a stamping plate defect detection system and method based on infrared imaging, including a stamping plate defect detection system based on infrared imaging, characterized in that it includes an acquisition and molding system and an imaging analysis system, wherein the acquisition and molding system includes a radar ultrasonic signal transmitting module, a radar ultrasonic signal receiving module, a radar parameter switching module, a radar ultrasonic signal verification module, a radar precision adjustment module, a radar data acquisition module, a radar data processing module, a radar data conversion module and a thermal imaging module connected in sequence, and the thermal imaging module is connected to an infrared laser transmitting module. The present invention can avoid the randomness of data caused by using only one detection method alone, and can also increase the accuracy of infrared imaging data, and increase the accuracy of visual inspection of stamping plates.

Description

Technical Field

[0001] The present invention relates to the technical field of plate defect detection, and in particular to a stamping plate defect detection system and method based on infrared imaging. Background Art

[0002] Stamping sheet defect detection is an important part of modern production, and the level of detection technology will directly affect the quality of stamping sheets. At present, most stamping sheets are still inspected by manual inspection, which not only has a large workload, but is also susceptible to subjective factors of quality inspectors, and cannot guarantee the accuracy and reliability of the inspection. The stamping sheet defect detection system based on machine vision has the advantages of fast detection speed, low cost and reliable detection results.

[0003] Machine vision inspection technology is that under certain lighting conditions (infrared, LED light source, etc.), the industrial camera collects the image of the stamping sheet to the edge server, and then uses image enhancement technology and machine vision technology to perform defect detection on the collected stamping sheet.

[0004] There are many types of defects in stamped sheets. Common defects in stamped sheets are mainly white spots, black spots, scratches, bruises and roller marks. These defects are accompanied by color difference of stamped sheets. Industrial cameras are used to obtain images of stamped sheets. Based on the images obtained by the camera, the defects of stamped sheets can be automatically detected using machine vision methods. The detection efficiency can be synchronized with the production rhythm.

[0005] In practical applications, a single nondestructive testing technology often cannot meet all testing needs. Different materials, structures and defect types may require a combination of multiple technologies for comprehensive evaluation. For example, in the aerospace field, multiple nondestructive testing technologies are often used to evaluate the integrity of aircraft structures. Ultrasonic testing can be used to detect internal cracks, foreign matter and corrosion; X-ray testing can find hidden welding defects; magnetic particle testing can be used to detect surface and near-surface cracks; penetrant testing is suitable for detecting small cracks and capillaries, etc. Through multiple testing technologies, the test results can be complemented, verified and confirmed to improve the reliability and accuracy of the test. Summary of the invention

[0006] In order to solve the technical problems existing in the prior art, the present invention provides a stamping plate defect detection system and method based on infrared imaging.

[0007] The present invention is implemented by the following technical scheme: a stamping sheet defect detection system based on infrared imaging, including an acquisition and molding system and an imaging analysis system, wherein the acquisition and molding system includes a radar ultrasonic signal transmitting module, a radar ultrasonic signal receiving module, a radar parameter switching module, a radar ultrasonic signal verification module, a radar precision adjustment module, a radar data acquisition module, a radar data processing module, a radar data conversion module and a thermal imaging module connected in sequence, and the thermal imaging module is connected to an infrared laser transmitting module;

[0008] The imaging analysis system includes an infrared imaging acquisition module, an infrared imaging information conversion module, an infrared imaging feature selection module, a feature matching module and an infrared imaging model training module which are connected in sequence. The thermal imaging module is connected to the infrared imaging acquisition module. The infrared imaging information conversion module is connected to an infrared imaging model input module which is connected to the infrared imaging model training module. The infrared imaging model training module is connected to an illumination calculation module. The illumination calculation module is connected to a temperature acquisition module, an illumination acquisition module, a temperature adjustment module and an illumination adjustment module.

[0009] As a further improvement of the above scheme, the radar ultrasonic signal transmitting module is used to transmit a radar ultrasonic signal for testing; the radar ultrasonic signal receiving module is used to receive the radar ultrasonic signal returned by the radar ultrasonic signal transmitting module after being reflected to the detection object; the radar parameter switching module is used to adjust the radar parameters when the received radar ultrasonic signal passes through different detection objects; the radar ultrasonic signal verification module is used to verify the radar ultrasonic signal received by the radar ultrasonic signal receiving module; the radar accuracy adjustment module is used to adjust the radar accuracy according to the verification result of the radar ultrasonic signal verification module; the radar data acquisition module is used to convert the radar ultrasonic signal verified by the radar ultrasonic signal verification module into radar data; the radar data processing module is used to process the radar data acquired by the radar data acquisition module; the radar data conversion module is used to convert the radar data processed by the radar data processing module into infrared data; the thermal imaging module is used to generate an infrared imaging image based on the infrared information emitted by the infrared laser transmitting module and the infrared information converted from the radar ultrasonic data; the infrared laser transmitting module is used to emit infrared laser for testing.

[0010] As a further improvement of the above scheme, the infrared imaging acquisition module is used to acquire infrared imaging information of the object to be tested that is currently undergoing infrared laser induced breakdown transmitted from the thermal imaging module; the infrared imaging information conversion module is used to convert the infrared imaging information at any temperature into its corresponding standard infrared imaging model when the temperature cannot be adjusted to the standard temperature; the infrared imaging feature selection module is used to select useful infrared imaging information collected in the infrared imaging acquisition module; the feature matching module is used to feature match the useful infrared imaging information selected in the infrared imaging feature selection module with the existing infrared imaging analysis model input by the infrared imaging model input module; the infrared imaging model training module is used to detect the standard stamping sheet placed on the inspection platform, and when the model training is terminated, output the model calibration performance parameters; the infrared imaging model input module is used to input the existing infrared imaging analysis model; the illumination calculation module is used to calculate the output optical power of the infrared laser emission module at the current temperature, and the temperature acquisition module is used to acquire the temperature of the object to be tested and the ambient temperature; the illumination acquisition module is used to acquire the ambient illumination and the output optical power of the infrared laser emission module; the temperature adjustment module is used to adjust the ambient temperature and the temperature of the object to be tested; the illumination adjustment module is used to acquire the ambient illumination and the output optical power of the infrared laser emission module.

[0011] As a further improvement of the above scheme, the temperature acquisition module includes an ambient temperature acquisition unit and a sample temperature acquisition unit connected to the illumination calculation module; the ambient temperature acquisition unit is used to collect the temperature of the environment outside the object to be tested; and the sample temperature acquisition unit is used to collect the temperature of the sample to be tested.

[0012] As a further improvement of the above scheme, the light collection module includes an ambient light collection unit and a sample light collection unit connected to the light calculation module; the ambient light collection unit uses the light from the environment outside the sample to be tested, and the sample light collection is used to collect the light from the infrared laser emission module that illuminates the sample to be tested.

[0013] A testing method for a stamping sheet metal defect detection system based on infrared imaging, comprising the following steps:

[0014] S1 infrared radar signal transmission:

[0015] The workpiece to be tested is placed at the testing position, the infrared laser is tested using the infrared laser transmitting module, and the radar ultrasonic signal is transmitted using the radar ultrasonic signal transmitting module;

[0016] S2 radar transmit signal processing:

[0017] The radar ultrasonic signal receiving module receives the radar ultrasonic signal, and the radar parameter switching module adjusts the radar parameters of the received radar ultrasonic signal. After that, the received radar ultrasonic signal is verified by the radar ultrasonic signal verification module. The radar ultrasonic signal that passes the verification is adjusted by the radar accuracy adjustment module to ensure the accuracy of the radar ultrasonic signal. For the radar ultrasonic signal that fails the verification, it returns to the radar parameter switching module to re-match the radar parameters.

[0018] The radar ultrasonic signal after precision adjustment is converted into radar data through the radar data acquisition module, and then the acquired radar data is processed by the radar data processing module, and the processed radar data is converted into infrared information through the radar data conversion module;

[0019] S3 infrared imaging image generation:

[0020] The infrared imaging module generates an infrared imaging image according to the infrared information of the infrared laser emission module and the infrared information converted from the radar ultrasonic data;

[0021] S4 inputs the infrared imaging model;

[0022] S5 performs information conversion, feature selection and matching on the acquired infrared imaging images and then inputs them into the infrared imaging model for training.

[0023] As a further improvement of the above solution, the working process of the infrared imaging information conversion module in step S5 is as follows:

[0024] a1. Read the infrared imaging data matrix A of different samples at room temperature, the sample concentration matrix B corresponding to A, the infrared imaging data matrix C of different samples at different temperatures, and the background intensity matrix D at different temperatures;

[0025] a2. Fit the infrared imaging data C at different temperatures into a curve C(D) along the background radiation D direction through functional data analysis, and fit the infrared imaging data A at normal temperature into a straight line A(D);

[0026] a3. Establish a nonlinear relationship between M infrared imaging curves C(D)1, C(D)2, ..., C(D)M at different temperatures and M normal temperature infrared imaging straight lines A(D)1, A(D)2, ..., A(D)M through functional model regression, and obtain a correction coefficient;

[0027] a4. Arrange the cycle results of W analysis lines in a matrix in sequence to form a nonlinear transformation model;

[0028] a5. Establish a least squares support vector machine regression model through the normal temperature data matrix A and its corresponding concentration matrix B;

[0029] a6. The near-infrared background of the test sample is used as the input of the conversion model to obtain the conversion coefficient; then, the infrared imaging intensity is multiplied by the conversion coefficient to obtain the converted infrared imaging; finally, the converted infrared imaging is used as the input of the regression model to obtain the infrared imaging data of the test sample.

[0030] As a further improvement of the above solution, the specific working process of the radar data processing module in step S2 is as follows:

[0031] b1. Call the preset number of sampling points for each channel;

[0032] b2. Read radar data;

[0033] b3. Convert radar data using bilinear interpolation algorithm;

[0034] b4. Generate color scale and complete rapid imaging, wherein the color scale is used to highlight abnormal areas in the generated image; wherein the direct wave and shallow strong reflection wave are cut off and then imaged again to highlight deep weak reflection wave;

[0035] b5. Use single target processing to process radar data by calling interference removal, path smoothing, digital filtering, attribute extraction, and offset;

[0036] b6, performing radar image denoising and target detection on the selected abnormal area, and then repeating step C5 to achieve fine processing of the selected abnormal area;

[0037] b7. Store the obtained radar image.

[0038] As a further improvement of the above solution, the step of training the infrared imaging model on the infrared imaging image in step S5 includes:

[0039] c1. Select data, randomly select a number of data from the infrared imaging model input module as a sample data group;

[0040] c2. Initialize each weight and bias, and set the adjustment parameters, accuracy requirement parameters and adjustment rate;

[0041] c3. For each sample in the training group, calculate the output of the measurement data of the standard stamping sheet and the sample data of the standard stamping sheet;

[0042] c4. Compare the elements in the predicted results with the elements in the sample results, and further calculate the error of the measured data;

[0043] c5. Calculate the increments of each weight and bias based on the result of step c4;

[0044] c6, update weights and biases;

[0045] c7, calculate the total error of the predicted data and compare it with the accuracy parameter to determine whether it is less than the accuracy control parameter. If not, return to step c3 to continue the operation;

[0046] c8. Adjustment is completed and the current weights and biases are saved. At this point, it can be considered that all parameters have reached a stable state and the training is completed.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] The present invention acquires infrared data information through an infrared laser emission module, and processes the acquired radar data through a radar data processing module to ensure the accuracy of the radar data, converts the processed radar data into infrared data through a radar data conversion module, generates an infrared imaging image based on the infrared information of the infrared laser emission module and the infrared information converted from the radar ultrasonic data through an infrared imaging module, and analyzes the infrared imaging image through an infrared imaging analysis system. This not only avoids the randomness of the data caused by using only one detection method, but also increases the accuracy of the infrared imaging data and the accuracy of the visual inspection of the stamped sheet. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A schematic structural diagram of a stamping sheet defect detection system based on infrared imaging provided by the present invention. DETAILED DESCRIPTION

[0050] The present invention is further described below in conjunction with the accompanying drawings and specific implementation methods. It should be noted that, under the premise of no conflict, the various embodiments or technical features described below can be arbitrarily combined to form a new embodiment.

[0051] Embodiment 1:

[0052] Please combine Figure 1 , a stamping sheet defect detection system based on infrared imaging in this embodiment includes an acquisition and molding system and an imaging analysis system, the acquisition and molding system includes a radar ultrasonic signal transmitting module, a radar ultrasonic signal receiving module, a radar parameter switching module, a radar ultrasonic signal verification module, a radar precision adjustment module, a radar data acquisition module, a radar data processing module, a radar data conversion module and a thermal imaging module connected in sequence, and the thermal imaging module is connected to an infrared laser transmitting module;

[0053] The imaging analysis system includes an infrared imaging acquisition module, an infrared imaging information conversion module, an infrared imaging feature selection module, a feature matching module and an infrared imaging model training module which are connected in sequence. The thermal imaging module is connected to the infrared imaging acquisition module. The infrared imaging information conversion module is connected to the infrared imaging model input module which is connected to the infrared imaging model training module. The infrared imaging model training module is connected to the illumination calculation module. The illumination calculation module is connected to the temperature acquisition module, the illumination acquisition module, the temperature regulation module and the illumination regulation module.

[0054] The radar ultrasonic signal transmitting module is used to transmit the radar ultrasonic signal for testing; the radar ultrasonic signal receiving module is used to receive the radar ultrasonic signal returned after being reflected by the radar ultrasonic signal transmitting module to the detection object; the radar parameter switching module is used to adjust the radar parameters when the received radar ultrasonic signal passes through different detection objects; the radar ultrasonic signal verification module is used to verify the radar ultrasonic signal received by the radar ultrasonic signal receiving module; the radar accuracy adjustment module is used to adjust the radar accuracy according to the verification result of the radar ultrasonic signal verification module; the radar data acquisition module is used to convert the radar ultrasonic signal verified by the radar ultrasonic signal verification module into radar data; the radar data processing module is used to process the radar data obtained by the radar data acquisition module; the radar data conversion module is used to convert the radar data processed by the radar data processing module into infrared data; the thermal imaging module is used to generate infrared imaging images according to the infrared information emitted by the infrared laser transmitting module and the infrared information converted from the radar ultrasonic data; the infrared laser transmitting module is used to emit infrared laser for testing.

[0055] The infrared imaging acquisition module is used to acquire the infrared imaging information of the object to be tested that is currently undergoing infrared laser induced breakdown transmitted from the thermal imaging module; the infrared imaging information conversion module is used to convert the infrared imaging information at any temperature into its corresponding standard infrared imaging model when the temperature cannot be adjusted to the standard temperature; the infrared imaging feature selection module is used to select the useful infrared imaging information collected in the infrared imaging acquisition module; the feature matching module is used to perform feature matching between the useful infrared imaging information selected in the infrared imaging feature selection module and the existing infrared imaging analysis model input by the infrared imaging model input module; the infrared imaging model training module is used to detect the standard stamping sheet placed on the inspection platform, and output the model calibration performance parameters when the model training is terminated; the infrared imaging model input module is used to input the existing infrared imaging analysis model; the illumination calculation module is used to calculate the output optical power of the infrared laser emission module at the current temperature, the temperature acquisition module is used to acquire the temperature of the object to be tested and the ambient temperature; the illumination acquisition module is used to acquire the ambient illumination and the output optical power of the infrared laser emission module; the temperature adjustment module is used to adjust the ambient temperature and the temperature of the object to be tested; the illumination adjustment module is used to acquire the ambient illumination and the output optical power of the infrared laser emission module.

[0056] The temperature acquisition module includes an environment temperature acquisition unit and a sample temperature acquisition unit connected to the illumination calculation module; the environment temperature acquisition unit is used to acquire the temperature of the environment outside the object to be tested; and the sample temperature acquisition unit is used to acquire the temperature of the sample to be tested.

[0057] The illumination acquisition module includes an ambient illumination acquisition unit and a sample illumination acquisition unit connected to the illumination calculation module; the ambient illumination acquisition unit uses the illumination of the environment outside the sample to be tested, and the sample illumination acquisition is used to collect the illumination of the infrared laser emission module that illuminates the sample to be tested.

[0058] Embodiment 2:

[0059] A testing method for a stamping sheet metal defect detection system based on infrared imaging comprises the following steps:

[0060] S1 infrared radar signal transmission:

[0061] The workpiece to be tested is placed at the testing position, the infrared laser is tested using the infrared laser transmitting module, and the radar ultrasonic signal is transmitted using the radar ultrasonic signal transmitting module;

[0062] S2 radar transmit signal processing:

[0063] The radar ultrasonic signal receiving module receives the radar ultrasonic signal, and the radar parameter switching module adjusts the radar parameters of the received radar ultrasonic signal. After that, the received radar ultrasonic signal is verified by the radar ultrasonic signal verification module. The radar ultrasonic signal that passes the verification is adjusted by the radar accuracy adjustment module to ensure the accuracy of the radar ultrasonic signal. For the radar ultrasonic signal that fails the verification, it returns to the radar parameter switching module to re-match the radar parameters.

[0064] The radar ultrasonic signal after precision adjustment is converted into radar data through the radar data acquisition module, and then the acquired radar data is processed by the radar data processing module, and the processed radar data is converted into infrared information through the radar data conversion module;

[0065] S3 infrared imaging image generation:

[0066] The infrared imaging module generates an infrared imaging image according to the infrared information of the infrared laser emission module and the infrared information converted from the radar ultrasonic data;

[0067] S4 inputs the infrared imaging model;

[0068] S5 performs information conversion, feature selection and matching on the acquired infrared imaging images and then inputs them into the infrared imaging model for training.

[0069] Embodiment 3:

[0070] The workflow of the infrared imaging information conversion module is as follows:

[0071] a1. Read the infrared imaging data matrix A of different samples at room temperature, the sample concentration matrix B corresponding to A, the infrared imaging data matrix C of different samples at different temperatures, and the background intensity matrix D at different temperatures;

[0072] a2. Fit the infrared imaging data C at different temperatures into a curve C(D) along the background radiation D direction through functional data analysis, and fit the infrared imaging data A at normal temperature into a straight line A(D);

[0073] a3. Establish a nonlinear relationship between M infrared imaging curves C(D)1, C(D)2, ..., C(D)M at different temperatures and M normal temperature infrared imaging straight lines A(D)1, A(D)2, ..., A(D)M through functional model regression, and obtain a correction coefficient;

[0074] a4. Arrange the cycle results of W analysis lines in a matrix in sequence to form a nonlinear transformation model;

[0075] a5. Establish a least squares support vector machine regression model through the normal temperature data matrix A and its corresponding concentration matrix B;

[0076] a6. The near-infrared background of the test sample is used as the input of the conversion model to obtain the conversion coefficient; then, the infrared imaging intensity is multiplied by the conversion coefficient to obtain the converted infrared imaging; finally, the converted infrared imaging is used as the input of the regression model to obtain the infrared imaging data of the test sample.

[0077] Embodiment 4:

[0078] The specific workflow of the radar data processing module is as follows:

[0079] b1. Call the preset number of sampling points for each channel;

[0080] b2. Read radar data;

[0081] b3. Convert radar data using bilinear interpolation algorithm;

[0082] b4. Generate color scale and complete rapid imaging, wherein the color scale is used to highlight abnormal areas in the generated image; wherein the direct wave and shallow strong reflection wave are cut off and then imaged again to highlight deep weak reflection wave;

[0083] b5. Use single target processing to process radar data by calling interference removal, path smoothing, digital filtering, attribute extraction, and offset;

[0084] b6, performing radar image denoising and target detection on the selected abnormal area, and then repeating step C5 to achieve fine processing of the selected abnormal area;

[0085] b7. Store the obtained radar image.

[0086] Embodiment 5:

[0087] The steps of training infrared imaging models on infrared imaging images include:

[0088] c1. Select data, randomly select a number of data from the infrared imaging model input module as a sample data group;

[0089] c2. Initialize each weight and bias, and set the adjustment parameters, accuracy requirement parameters and adjustment rate;

[0090] c3. For each sample in the training group, calculate the output of the measurement data of the standard stamping sheet and the sample data of the standard stamping sheet;

[0091] c4. Compare the elements in the predicted results with the elements in the sample results, and further calculate the error of the measured data;

[0092] c5. Calculate the increments of each weight and bias based on the result of step c4;

[0093] c6, update weights and biases;

[0094] c7, calculate the total error of the predicted data and compare it with the accuracy parameter to determine whether it is less than the accuracy control parameter. If not, return to step c3 to continue the operation;

[0095] c8. Adjustment is completed and the current weights and biases are saved. At this point, it can be considered that all parameters have reached a stable state and the training is completed.

[0096] Embodiment 6:

[0097] The radar data conversion module uses data fusion to mix data, which specifically includes the following steps:

[0098] d1. Assume that the error estimates of the infrared sensor and radar for the same target state at a certain moment are independent of each other and use event H0 to represent the infrared sensor data. and radar infrared data Coming from the same goal;

[0099] d2, event H1 indicates and From different goals;

[0100] d3. The real status of the target at this time is and The target state estimation error is and The covariance matrix of the statistic is:

[0101] in, is the infrared sensor’s estimated error covariance of the infrared data of environment i at time t, is the error covariance of the infrared data estimation of the radar infrared data for the environment j at time t;

[0102] d4, and then the test statistic is:

[0103]

[0104] Among them, α ij (t) obeys the x2 distribution with nx degrees of freedom, where nx is the dimension of the state estimation vector. ij When (t) is less than the threshold value, it means that the infrared data of environment i and the infrared data of environment j are related, otherwise they are not related;

[0105] d5. After the local infrared data information of the infrared sensor and the radar are associated, the successfully associated infrared data needs to be fused to obtain the final infrared data information. The fused result is:

[0106]

[0107] Among them, x1 and P1 are the filter value and error covariance matrix of the infrared sensor, x2 and P2 are the filter value and error covariance matrix of the radar, and xF and PF are the fused state and error covariance matrix, respectively.

[0108] The above-mentioned setting is adopted to obtain infrared data information through the infrared laser transmitting module, transmit the radar ultrasonic signal through the radar ultrasonic signal transmitting module, receive the radar ultrasonic signal through the radar ultrasonic signal receiving module, and adjust the radar parameters through the radar parameter switching module when the received radar ultrasonic signal passes through different layers. In this way, different radar ultrasonic signals can be automatically switched when the radar ultrasonic signal passes through different layers, so as to ensure the optimal matching result between the radar ultrasonic signal and the current layer, and avoid errors in the radar detection results due to poor matching results between the radar ultrasonic signal and the current layer. The received radar ultrasonic signal is verified by the radar ultrasonic signal verification module, and the qualified radar ultrasonic signal is verified. Perform radar accuracy adjustment to ensure the accuracy of the radar ultrasonic signal. For radar ultrasonic signals that fail the verification, return to the radar parameter switching module to re-match the radar parameters to ensure the accuracy of the radar ultrasonic signal parameters. The radar ultrasonic signal after the accuracy adjustment is converted into radar data through the radar data acquisition module. The acquired radar data is processed through the radar data processing module to ensure the accuracy of the radar data. The processed radar data is converted into infrared data through the radar data conversion module. The infrared imaging module generates an infrared imaging image according to the infrared information of the infrared laser emission module and the infrared information converted from the radar ultrasonic data. The infrared imaging image is analyzed by the infrared imaging analysis system.

[0109] Working principle:

[0110] The staff inputs the existing infrared laser analysis model through the infrared laser model input module, places the standard stamping sheet on the sample placement platform, starts the infrared laser emission module to perform infrared laser induced breakdown on the sample to be tested, collects the infrared laser information of the object to be tested that is currently undergoing infrared laser induced breakdown through the infrared laser acquisition module, selects the useful infrared laser information collected in the infrared laser acquisition module through the infrared laser feature selection module, and performs feature matching between the useful infrared laser information selected in the infrared laser feature selection module and the existing infrared laser analysis model input by the infrared laser model input module through the feature matching module. For infrared laser models with the same feature matching, the current corresponding environmental information can be known. At this time, feature matching can be used to achieve common identification of multiple elements. The temperature of the object to be tested and the ambient temperature are collected through the temperature acquisition module, and the ambient temperature and the temperature of the object to be tested are adjusted through the temperature adjustment module. This setting can ensure that the detection temperature of the object is consistent with the establishment temperature of the existing infrared laser analysis model in the system, avoiding data abnormalities due to changes in the detection temperature; the ambient light and the output optical power of the infrared laser emission module are collected through the light acquisition module, and the light adjustment module is used to adjust the ambient temperature. The ambient light and the output optical power of the infrared laser emitting module are set to ensure that the ambient light of the object and the output optical power of the infrared laser emitting module are consistent with the establishment conditions of the existing infrared laser analysis model in the system, avoiding data abnormalities due to changes in the detection environment. However, as the temperature of the infrared laser emitting module increases, the threshold current of the infrared laser increases and the output power decreases. Therefore, it is necessary to calculate the output optical power of the infrared laser emitting module at the current temperature through the illumination calculation module, so that the illumination adjustment module can adjust the illumination intensity of the detection environment. By obtaining the output optical power of the current infrared laser emitting module, the working state of the infrared laser emitting module can also be monitored. When the temperature adjustment module cannot adjust the ambient temperature to be consistent with the establishment temperature of the existing infrared laser analysis model in the system, the infrared laser information conversion module can be used to convert the infrared laser information at any temperature to its corresponding standard infrared laser model when the temperature cannot be adjusted to the standard temperature. In order to ensure the working accuracy of the detection and analysis system, the infrared laser model training module can be used to place a standard stamping sheet on the inspection platform for detection, and when the model training is terminated, the model calibration performance parameters are output to achieve system training.

[0111] The present invention acquires infrared data information through an infrared laser emission module, and processes the acquired radar data through a radar data processing module to ensure the accuracy of the radar data, converts the processed radar data into infrared data through a radar data conversion module, generates an infrared imaging image based on the infrared information of the infrared laser emission module and the infrared information converted from the radar ultrasonic data through an infrared imaging module, and analyzes the infrared imaging image through an infrared imaging analysis system. This not only avoids the randomness of the data caused by using only one detection method, but also increases the accuracy of the infrared imaging data and the accuracy of the visual inspection of the stamped sheet.

[0112] The above-mentioned embodiments are only preferred embodiments of the present invention and cannot be used to limit the scope of protection of the present invention. Any non-substantial changes and substitutions made by technicians in this field on the basis of the present invention shall fall within the scope of protection required by the present invention.

Claims

1. A testing method for a stamping sheet defect detection system based on infrared imaging, characterized in that: The detection system includes an acquisition and molding system and an imaging analysis system. The acquisition and molding system includes a radar ultrasonic signal transmitting module, a radar ultrasonic signal receiving module, a radar parameter switching module, a radar ultrasonic signal verification module, a radar precision adjustment module, a radar data acquisition module, a radar data processing module, a radar data conversion module and a thermal imaging module connected in sequence. The thermal imaging module is connected to an infrared laser transmitting module; The imaging analysis system comprises an infrared imaging acquisition module, an infrared imaging information conversion module, an infrared imaging feature selection module, a feature matching module and an infrared imaging model training module which are connected in sequence, the thermal imaging module is connected to the infrared imaging acquisition module, the infrared imaging information conversion module is connected to an infrared imaging model input module which is connected to the infrared imaging model training module, the infrared imaging model training module is connected to an illumination calculation module, the illumination calculation module is connected to a temperature acquisition module, an illumination acquisition module, a temperature adjustment module and an illumination adjustment module; The infrared imaging acquisition module is used to acquire infrared imaging information of the object to be tested that is currently undergoing infrared laser induced breakdown and transmitted from the thermal imaging module; the infrared imaging information conversion module is used to convert the infrared imaging information at any temperature into its corresponding standard infrared imaging model when the temperature cannot be adjusted to the standard temperature; the infrared imaging feature selection module is used to select useful infrared imaging information collected in the infrared imaging acquisition module; the feature matching module is used to perform feature matching between the useful infrared imaging information selected in the infrared imaging feature selection module and the existing infrared imaging analysis model input by the infrared imaging model input module; the infrared imaging model training module is used to detect the standard stamping sheet placed on the inspection platform, and output the model calibration performance parameters when the model training is terminated; the infrared imaging model input module is used to input the existing infrared imaging analysis model; the illumination calculation module is used to calculate the output optical power of the infrared laser emission module at the current temperature, and the temperature acquisition module is used to acquire the temperature of the object to be tested and the ambient temperature; the illumination acquisition module is used to acquire the ambient illumination and the output optical power of the infrared laser emission module; the temperature adjustment module is used to adjust the ambient temperature and the temperature of the object to be tested; the illumination adjustment module is used to acquire the ambient illumination and the output optical power of the infrared laser emission module; The test method includes the following steps: S1 infrared radar signal transmission: The workpiece to be tested is placed at the testing position, and the infrared laser transmitting module is used to transmit the infrared laser for testing, and the radar ultrasonic signal transmitting module is used to transmit the radar ultrasonic signal for testing; S2 radar transmit signal processing: The radar ultrasonic signal receiving module receives the radar ultrasonic signal, and the radar parameter switching module adjusts the radar parameters of the received radar ultrasonic signal. After that, the received radar ultrasonic signal is verified by the radar ultrasonic signal verification module. The radar ultrasonic signal that passes the verification is adjusted by the radar accuracy adjustment module to ensure the accuracy of the radar ultrasonic signal. For the radar ultrasonic signal that fails the verification, it returns to the radar parameter switching module to re-match the radar parameters. The radar ultrasonic signal after precision adjustment is converted into radar data through the radar data acquisition module, and then the acquired radar data is processed by the radar data processing module, and the processed radar data is converted into infrared information through the radar data conversion module; S3 infrared imaging image generation: The infrared imaging module generates an infrared imaging image according to the infrared information of the infrared laser emission module and the infrared information converted from the radar ultrasonic data; S4 inputs the infrared imaging model; S5 performs information conversion, feature selection and matching on the collected infrared imaging images and then inputs them into the infrared imaging model for training; The workflow of the infrared imaging information conversion module in step S5 is as follows: a1. Read the infrared imaging data matrix A of different samples at room temperature, the sample concentration matrix B corresponding to A, the infrared imaging data matrix C of different samples at different temperatures, and the background intensity matrix D at different temperatures; a2. Through functional data analysis, the infrared imaging data C at different temperatures are fitted into a curve C(D) along the background radiation D direction, and the infrared imaging data A at normal temperature is fitted into a straight line A(D); a3. Establish the nonlinear relationship between M infrared imaging curves of different temperatures C(D)1, C(D)2, ..., C(D)M and M normal temperature infrared imaging straight lines A(D)1, A(D)2, ..., A(D)M through functional model regression, and obtain the correction coefficient; a4. Arrange the cycle results of W analysis lines in a matrix to form a nonlinear transformation model; a5. Establish a least squares support vector machine regression model through the normal temperature data matrix A and its corresponding concentration matrix B; a6. The near-infrared background of the test sample is used as the input of the conversion model to obtain the conversion coefficient; then, the infrared imaging intensity is multiplied by the conversion coefficient to obtain the converted infrared imaging; finally, the converted infrared imaging is used as the input of the regression model to obtain the infrared imaging data of the test sample.

2. The testing method according to claim 1, characterized in that: The radar ultrasonic signal transmitting module is used to transmit the radar ultrasonic signal for testing; the radar ultrasonic signal receiving module is used to receive the radar ultrasonic signal returned after being reflected by the radar ultrasonic signal transmitting module to the detection object; the radar parameter switching module is used to adjust the radar parameters when the received radar ultrasonic signal passes through different detection objects; the radar ultrasonic signal verification module is used to verify the radar ultrasonic signal received by the radar ultrasonic signal receiving module; the radar accuracy adjustment module is used to adjust the radar accuracy according to the verification result of the radar ultrasonic signal verification module; the radar data acquisition module is used to convert the radar ultrasonic signal verified by the radar ultrasonic signal verification module into radar data; the radar data processing module is used to process the radar data obtained by the radar data acquisition module; the radar data conversion module is used to convert the radar data processed by the radar data processing module into infrared data; the thermal imaging module is used to generate an infrared imaging image according to the infrared information emitted by the infrared laser transmitting module and the infrared information converted from the radar ultrasonic data; the infrared laser transmitting module is used to emit infrared laser for testing.

3. The testing method according to claim 1, characterized in that: The temperature acquisition module includes an environment temperature acquisition unit and a sample temperature acquisition unit connected to the illumination calculation module; the environment temperature acquisition unit is used to acquire the temperature of the environment outside the object to be tested; and the sample temperature acquisition unit is used to acquire the temperature of the sample to be tested.

4. The testing method according to claim 1, characterized in that: The illumination acquisition module includes an ambient illumination acquisition unit and a sample illumination acquisition unit connected to the illumination calculation module; the ambient illumination acquisition unit uses the illumination of the environment outside the sample to be tested, and the sample illumination acquisition is used to collect the illumination of the infrared laser emission module that illuminates the sample to be tested.

5. The testing method according to claim 1, characterized in that: The specific working process of the radar data processing module in step S2 is as follows: b1. Call the preset number of sampling points for each channel; b2. Read radar data; b3. Convert radar data using bilinear interpolation algorithm; b4. Generate color scale and complete rapid imaging, wherein the color scale is used to highlight abnormal areas in the generated image; wherein the direct wave and shallow strong reflection wave are cut off and then imaged again to highlight deep weak reflection wave; b5. Use single target processing to process radar data by calling interference removal, path smoothing, digital filtering, attribute extraction, and offset; b6, performing radar image denoising and target detection on the selected abnormal area, and then repeating step C5 to achieve fine processing of the selected abnormal area; b7. Store the obtained radar image.

6. The testing method according to claim 1, characterized in that: The step of training the infrared imaging model on the infrared imaging image in step S5 includes: c1. Select data, randomly select a number of data from the infrared imaging model input module as a sample data group; c2. Initialize each weight and bias, and set the adjustment parameters, accuracy requirement parameters and adjustment rate; c3. For each sample in the training group, calculate the output of the measurement data of the standard stamping sheet and the sample data of the standard stamping sheet; c4. Compare the elements in the predicted results with the elements in the sample results, and further calculate the error of the measured data; c5. Calculate the increments of each weight and bias based on the result of step c4; c6, update weights and biases; c7, calculate the total error of the predicted data and compare it with the accuracy parameter to determine whether it is less than the accuracy control parameter. If not, return to step c3 to continue the operation; c8. Adjustment is completed and the current weights and biases are saved. At this point, it can be considered that all parameters have reached a stable state and the training is completed.

Images