Equipment damage detection method based on ultrasonic infrared thermal imaging time sequence self-weighting

Abstract

The invention discloses an equipment damage detection method based on ultrasonic infrared thermal imaging time sequence self-weighting. The method comprises the following steps: 1, acquiring an ultrasonic infrared thermal imaging sequence of equipment to be detected and cutting to obtain an infrared thermal map; 2, processing the infrared thermal image by using an entropy rate superpixel segmentation method to obtain superpixel blocks; 3, acquiring a pixel anchor point similarity matrix of each infrared thermal image; 4, on the basis of the pixel anchor point similarity matrix, setting a fusion similarity matrix and an adaptive weighted weight of each infrared thermal image, and constructing an ultrasonic infrared thermal imaging time sequence self-weighted fusion model; 5, performing optimization solution based on the ultrasonic infrared thermal imaging time sequence self-weighted fusion model to obtain an optimal fusion similarity matrix; and 6, obtaining a damaged region by using spectral clustering based on the optimal fusion similarity matrix. The method is simple in step, damage area segmentation is achieved based on the fusion similarity matrix, manual information labeling is not needed, the weight of the time sequence diagram is dynamically adjusted, and unsupervised equipment damage detection is achieved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of equipment damage image detection, and particularly relates to an equipment damage detection method based on ultrasonic infrared thermal imaging time sequence self-weighting. BACKGROUND

[0002] In nondestructive testing technology, although traditional ultrasonic testing can provide high-resolution internal defect information, it needs coupling agent and has slow detection speed; X-ray testing can penetrate materials and generate two-dimensional images of internal structures, but it is expensive and has radiation safety problems. Ultrasonic infrared thermal imaging has long been widely concerned due to its fast detection speed, wide detection area and strong adaptability. This technology uses ultrasonic excitation to induce frictional heat in the internal damage area of the detected object, and real-time acquisition by an infrared thermal imager can realize detection, which has unique advantages. Using ultrasonic infrared thermal imaging for nondestructive testing of equipment can effectively solve the limitations of existing traditional methods in the above aspects.

[0003] How to accurately extract the damage area in ultrasonic infrared thermal imaging is a key problem. However, the damage edge blur caused by heat diffusion in ultrasonic infrared thermal imaging is easy to cause misjudgment. In addition, due to the difference in heat diffusion of different equipment, it is difficult to manually weigh the specific time period of the warming, constant temperature and cooling process and select the input image frame. More importantly, the current method often needs to pre-label the damage area and relies on a large amount of labeled data for training.

[0004] Therefore, at present, there is a lack of a kind of design reasonable based on ultrasonic infrared thermal imaging time sequence self-weighting equipment damage detection method, realize ultrasonic infrared thermal imaging sequence self-adaptive weighting, obtain fusion similarity matrix and get damage area, without manual annotation information, realize unsupervised equipment damage detection. SUMMARY

[0005] The technical problem to be solved by the present application is to solve the above-mentioned deficiencies in the prior art, to provide an equipment damage detection method based on ultrasonic infrared thermal imaging time sequence self-weighting, which has simple method steps, reasonable design, realizes self-adaptive weighting of ultrasonic infrared thermal imaging sequence, obtains fusion similarity matrix and gets damage area, without manual annotation information, realizes unsupervised equipment damage detection.

[0006] To solve the above technical problems, the technical scheme adopted by the present application is: an equipment damage detection method based on ultrasonic infrared thermal imaging time sequence self-weighting, which comprises the following steps: Step one, collect the ultrasonic infrared thermal imaging sequence of the equipment to be tested and cut to obtain infrared thermal images; Step two, use the entropy rate superpixel segmentation method to process the infrared thermal images to obtain superpixel blocks; Step three, obtain the pixel anchor point similarity matrix of each infrared thermal image; Step four, based on the pixel anchor point similarity matrix, set the fusion similarity matrix and the adaptive weighting weight of each infrared thermal image, and construct the ultrasonic infrared thermal imaging time sequence self-weighted fusion model; Step five, based on the ultrasonic infrared thermal imaging time sequence self-weighted fusion model, the optimal fusion similarity matrix is obtained by optimization solution; Step six, based on the optimal fusion similarity matrix, the damage area is obtained by using spectral clustering.

[0007] The above-mentioned device damage detection method based on ultrasonic infrared thermal imaging time sequence self-weighting, further, step one, the specific process is as follows: Step 101, using an ultrasonic infrared thermal image detection system to collect infrared thermal images of the device to be tested, and obtaining an ultrasonic infrared thermal imaging sequence; Step 102, using a computer to take the t-th frame of original infrared thermal image from the ultrasonic infrared thermal imaging sequence, and cutting the edge of the t-th frame of original infrared thermal image, to obtain the t-th frame of cut infrared thermal image, wherein the value range of t is 1~N, and N represents the length of the ultrasonic infrared thermal imaging sequence; Step 103, repeat step 102 multiple times to complete the cutting of the N-th frame of original infrared thermal image, and obtain the N-th frame of cut infrared thermal image; Step 104, the first frame of cut infrared thermal image to the N-th frame of cut infrared thermal image is recorded as N infrared thermal images.

[0008] The above-mentioned device damage detection method based on ultrasonic infrared thermal imaging time sequence self-weighting, further, step two, the specific process is as follows: Step 201, combining N infrared thermal images according to channel array to obtain a high-dimensional infrared thermal image of w×h×N; wherein w represents the width of the cut infrared thermal image, and h represents the height of the cut infrared thermal image; Step 202, converting the high-dimensional infrared thermal image of w×h×N into a two-dimensional matrix; wherein the number of rows of the two-dimensional matrix is w×h, and the number of columns of the two-dimensional matrix is N; Step 203, using a computer to reduce the dimension of the two-dimensional matrix by principal component analysis, and taking the first principal component as the reduced two-dimensional matrix; Step 204, using a computer to convert the reduced two-dimensional matrix into a two-dimensional gray image; Step 205, using a computer to segment the two-dimensional gray image according to the set number of superpixel blocks I' by using the entropy rate superpixel segmentation method, to obtain I' superpixel blocks; wherein I' is a positive integer.

[0009] The above-mentioned device damage detection method based on ultrasonic infrared thermal imaging time sequence self-weighting, further, step three, the specific process is as follows: Step 301, using a computer to obtain the anchor point of the i-th superpixel block on the t-th infrared thermal image according to ; wherein, represents the pixel value of the e-th pixel point in the i-th superpixel block on the t-th infrared thermal image, and the value range of e is , the value range of i is 1~I', represents the total number of pixel points in the i-th superpixel block; Step 302, using a computer to obtain the similarity value between the j-th pixel point and the anchor point of the i-th superpixel block on the t-th infrared thermal image according to ; wherein, represents the pixel value of the j-th pixel point on the t-th infrared thermal image, and the value range of j is 1~w×h; Step 303, repeatedly step 302 multiple times to obtain the similarity value between each pixel point and each anchor point on the t-th infrared thermal image, and combine the similarity value between each pixel point and each anchor point into a pixel-anchor point similarity matrix of the t-th infrared thermal image .

[0010] The above-mentioned device damage detection method based on ultrasonic infrared thermal imaging time sequence self-weighting, further, step four, the specific process is as follows: using a computer to set the fusion similarity matrix corresponding to N infrared thermal images as , input the pixel-anchor point similarity matrix of the t-th infrared thermal image , and construct an ultrasonic infrared thermal imaging time sequence self-weighting fusion model ; wherein, min represents the minimum value, s.t. represents the constraint condition, represents the square of F norm, represents the hyperparameter, represents the natural logarithm function, the number of rows of which is w×h, and the number of columns of which is I'; is a first constant matrix, and each element in which is 1, and the size of which is I'×1; is a second constant matrix, and each element in which is 1, and the number of rows of which is w×h, and the number of columns of which is 1; represents the adaptive weighting weight during fusion of the t-th infrared thermal image.

[0011] The above-mentioned device damage detection method based on ultrasonic infrared thermal imaging time sequence self-weighting, further, step five, the specific process is as follows: ​​​Step 501, adopt a computer to make the initial value of the adaptive weighting weight in the fusion of each infrared thermal image be ; Step 502, input the pixel anchor point similarity matrix of the tth infrared thermal image , and let Substitute into the ultrasonic infrared thermal imaging time sequence self-weighted fusion model to obtain the converted ultrasonic infrared thermal imaging time sequence self-weighted fusion model ; Step 503, solve the converted ultrasonic infrared thermal imaging time sequence self-weighted fusion model to obtain the optimal fusion similarity matrix after the first iteration ; Step 504, use a computer to obtain the adaptive weighting weight of the tth infrared thermal image fusion after the first iteration according to ; ; Step 505, according to the method of steps 502 to 504, let , substitute the adaptive weighting weight of the tth infrared thermal image fusion after the first iteration , and solve to obtain the optimal fusion similarity matrix after the second iteration and the adaptive weighting weight of the tth infrared thermal image fusion after the second iteration ; Step 506, repeatedly step 505 multiple times, let , substitute the adaptive weighting weight of the tth infrared thermal image fusion after the k-1th iteration , and solve to obtain the optimal fusion similarity matrix after the kth iteration and the adaptive weighting weight of the tth infrared thermal image fusion after the kth iteration ; wherein k is a positive integer greater than 2; Step 507, the difference between the minimum value of the objective function of the kth iteration and the minimum value of the objective function of the k-1th iteration is less than 0.01, then the iteration is ended, and the optimal fusion similarity matrix after the kth iteration is recorded as the optimal fusion similarity matrix.

[0012] The above-mentioned device damage detection method based on ultrasonic infrared thermal imaging time sequence self-weighting further comprises step six, and the specific process is as follows: Step 601, use a computer to set the category number to 2, and use the spectral clustering method to cluster process the optimal fusion similarity matrix to obtain pixel points with a category label of 1 and pixel points with a category label of 2; In step 602, the computer is used to obtain the pixel mean of the pixel points with the category label 1 and the pixel mean of the pixel points with the category label 2, and the pixel region corresponding to the category label with the larger pixel mean is the damage region.

[0013] Compared with the prior art, the present application has the following advantages: 1. The method of the present application has simple steps and reasonable design. Firstly, an infrared thermal image is obtained by cutting an ultrasonic infrared thermal imaging sequence. Secondly, an entropy rate superpixel segmentation method is used to process the infrared thermal image to obtain a superpixel block. Then, a pixel anchor point similarity matrix of each infrared thermal image is obtained. After that, an ultrasonic infrared thermal imaging time sequence self-weighted fusion model is constructed to obtain an optimal fusion similarity matrix. Finally, damage region segmentation is realized based on the fusion similarity matrix classification without manual annotation information, and the time sequence graph weight is dynamically adjusted to obtain an accurate damage region in an unsupervised manner.

[0014] 2. The present application first uses an entropy rate superpixel segmentation method to process the infrared thermal image to obtain a superpixel block, which facilitates subsequent classification of each pixel anchor point similarity matrix based on a unified superpixel block, thereby facilitating subsequent fusion.

[0015] 3. The present application sets a fusion affinity matrix, which is obtained by considering the pixel anchor point similarity matrix of each infrared thermal image. On the one hand, it reflects the damage detection of the ultrasonic infrared thermal imaging sequence. On the other hand, it can avoid the interference of unstable time sequence graphs to accurately reflect the stable region of the damage. In addition, the adaptive weighting weight of each infrared thermal image is considered, and the weight is adaptively updated to suppress the negative effects caused by the instability of the early image frames of heat diffusion, thereby avoiding the subjective bias caused by the heat diffusion difference of different damage conditions when manually intercepting the time sequence, and effectively enhancing the accuracy of damage segmentation.

[0016] 4. In the self-weighted fusion model of the present application, a reasonable weight is generated by the pixel anchor point similarity matrix of each infrared thermal image and the quantitative difference of all frame fusion results, and the weight is further scaled by a designed function to improve the importance of stable frames and make unstable frames with large differences have a smaller weight after scaling, thereby improving the robustness.

[0017] In summary, the method of the present application has simple steps and reasonable design. Based on the pixel anchor point similarity matrix and the fusion similarity matrix of each infrared thermal image, and considering the adaptive weighting weight of each infrared thermal image, an ultrasonic infrared thermal imaging time sequence self-weighted fusion model is constructed to obtain an optimal fusion similarity matrix and realize damage region segmentation without manual annotation information, and the time sequence graph weight is dynamically adjusted to realize unsupervised device damage detection.

[0018] The technical solutions of the present application will be further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 Flow chart of the method of the present application.

[0020] Figure 2 Schematic diagram of damage detection after 20J energy impact of the present application.

[0021] Figure 3 Schematic diagram of damage detection after 40J energy impact of the present application. DETAILED DESCRIPTION

[0022] As shown in Figure 1 , a device damage detection method based on ultrasonic infrared thermal imaging time sequence self-weighting of the present application, the method comprising the following steps: Step one, collect the ultrasonic infrared thermal imaging sequence of the device to be tested and crop to obtain infrared thermal images; Step two, use the entropy rate superpixel segmentation method to process the infrared thermal images to obtain superpixel blocks; Step three, obtain the pixel anchor point similarity matrix of each infrared thermal image; Step four, based on the pixel anchor point similarity matrix, set the fusion similarity matrix and the adaptive weighting weight of each infrared thermal image, and construct an ultrasonic infrared thermal imaging time sequence self-weighting fusion model; Step five, based on the ultrasonic infrared thermal imaging time sequence self-weighting fusion model, optimize and solve to obtain the optimal fusion similarity matrix; Step six, based on the optimal fusion similarity matrix, use spectral clustering to obtain the damage area.

[0023] In this embodiment, step one has the following specific process: Step 101, use an ultrasonic infrared thermal imaging detection system to collect infrared thermal images of the device to be tested to obtain an ultrasonic infrared thermal imaging sequence; Step 102, use a computer to take the tthframe of the original infrared thermal image from the ultrasonic infrared thermal imaging sequence, and crop the edges of the tthframe of the original infrared thermal image to obtain the tthframe of the cropped infrared thermal image, wherein the value of t ranges from 1 to N, and N represents the length of the ultrasonic infrared thermal imaging sequence; Step 103, repeat step 102 multiple times to complete the cropping of the Nthframe of the original infrared thermal image to obtain the Nthframe of the cropped infrared thermal image; Step 104, label the 1stframe of the cropped infrared thermal image to the Nthframe of the cropped infrared thermal image as N infrared thermal images.

[0024] In this embodiment, step two has the following specific process: Step 201, combine the N infrared thermal images according to the channel array to obtain a high-dimensional infrared thermal image of w x h x N; wherein w represents the width of the cropped infrared thermal image, and h represents the height of the cropped infrared thermal image; Step 202: Convert the w×h×N high-dimensional infrared thermal image into a two-dimensional matrix; where the number of rows in the two-dimensional matrix is ​​w×h and the number of columns in the two-dimensional matrix is ​​N; Step 203: Use a computer to perform dimensionality reduction on the two-dimensional matrix using principal component analysis, and use the first principal component as the dimensionality-reduced two-dimensional matrix; Step 204: Use a computer to convert the dimensionality-reduced two-dimensional matrix into a two-dimensional grayscale image; Step 205: Using a computer, the entropy rate superpixel segmentation method is used to segment the two-dimensional grayscale image according to a set number of superpixel blocks I', resulting in I' superpixel blocks; where I' is a positive integer.

[0025] In this embodiment, step three is as follows: Step 301: Using a computer to... The anchor point of the i-th superpixel block on the t-th infrared thermal image is obtained. ;in, This represents the pixel value of the e-th pixel in the i-th superpixel block of the t-th infrared thermal image, where the value of e ranges from 0 to 1. The value of i ranges from 1 to I'. This represents the total number of pixels in the i-th superpixel block; Step 302: Using a computer to... The anchor point between the j-th pixel and the i-th superpixel block on the t-th infrared thermal image is obtained. Similarity value between ;in, This represents the pixel value of the j-th pixel on the t-th infrared thermal image, where the value of j ranges from 1 to w×h; Step 303: Repeat step 302 multiple times to obtain the similarity value between each pixel and each anchor point in the t-th infrared thermal image, and combine the similarity values ​​between each pixel and each anchor point to form the pixel-anchor-similarity matrix of the t-th infrared thermal image. .

[0026] In this embodiment, step four is as follows: The computer sets the fusion similarity matrix corresponding to N infrared thermal images as follows: Input the pixel anchor point similarity matrix of the t-th infrared thermal image. Construct a time-series self-weighted fusion model of ultrasonic infrared thermal imaging Where min represents the minimum value and st represents the constraint condition. Denotes the square of the F-norm. Indicates hyperparameters, Represents the natural logarithm function. The number of rows is w×h, and the number of columns is I'; It is the first constant matrix, and Each element in the array is 1, and Its size is I' × 1; It is the second constant matrix, and Each element in the array is 1, and The number of rows is w×h, and the number of columns is 1; This represents the adaptive weighting weight when fusing the t-th infrared thermal image.

[0027] In this embodiment, step five is as follows: Step 501: Use a computer to initialize the adaptive weighting weights for fusing the various infrared thermal images to the following values: ; Step 502: Input the pixel anchor point similarity matrix of the t-th infrared thermal image. and order Substituting the data into the time-series self-weighted fusion model of ultrasound infrared thermal imaging, we obtain the transformed time-series self-weighted fusion model of ultrasound infrared thermal imaging. ; Step 503: Solve the converted ultrasonic infrared thermal imaging time-series self-weighted fusion model to obtain the optimal fusion similarity matrix after the first iteration. ; Step 504: Using a computer to... The adaptive weighting weights for the t-th infrared thermal image fusion after the first iteration are obtained. ; Step 505: Following the methods in steps 502 to 504, let The adaptive weighting weights are used when fusing the t-th infrared thermal image after the first iteration. Substituting the values ​​and solving the problem, we obtain the optimal fusion similarity matrix after the second iteration. Adaptive weighting weights when fusing with the t-th infrared thermal image after the second iteration ; Step 506: Repeat step 505 multiple times, let The adaptive weighting weights used when fusing the t-th infrared thermal image after the (k-1)-th iteration Substituting the values ​​and solving the problem, we obtain the optimal fusion similarity matrix after the k-th iteration. Adaptive weighting weights when fusing with the t-th infrared thermal image after the k-th iteration Where k is a positive integer greater than 2; Step 507: If the difference between the minimum value of the objective function in the k-th iteration and the minimum value of the objective function in the (k-1)-th iteration is less than 0.01, the iteration ends, and the optimal fusion similarity matrix after the k-th iteration is calculated. Let this be denoted as the optimal fusion similarity matrix.

[0028] In this embodiment, step six, the specific process is as follows: Step 601, using a computer to set the number of categories to 2, and using a spectral clustering method to cluster the optimal fusion similarity matrix to obtain pixel points with a category label of 1 and pixel points with a category label of 2; Step 602, using a computer to obtain the pixel mean of the pixel points with a category label of 1 and the pixel mean of the pixel points with a category label of 2, and then the pixel region corresponding to the category label with the larger pixel mean is the damage region.

[0029] In this embodiment, in specific implementation, the device to be tested is specifically: A carbon fiber reinforced polyether ether ketone composite sample plate, the length of the sample plate is 150mmx100mmx3mm; a drop hammer impact testing machine of DI'T183E type is used to perform vertical impact tests on the sample plate with energies of 20J and 40J, respectively, to obtain an impact damaged sample plate.

[0030] In this embodiment, in specific implementation, the infrared thermograph acquisition is performed on the impact damaged sample plate using an ultrasonic infrared thermograph detection system.

[0031] In this embodiment, the ultrasonic infrared thermograph detection system includes an infrared thermal imager, an ultrasonic excitation source, and a computer, the infrared thermal imager is a VCHD research 680 infrared thermal imager manufactured by InfraTec Company of Germany; the ultrasonic excitation source is a UST-200 ultrasonic gun produced by Nanjing Novell Photoelectric System Co., Ltd., and the acquired ultrasonic infrared thermal imaging sequence is sent to the computer for processing.

[0032] In this embodiment, N represents the length of the ultrasonic infrared thermal imaging sequence, and N=800; The value of I' is 145.

[0033] In this embodiment, It is indicated that each element value in the matrix is greater than or equal to zero.

[0034] In this embodiment, it should be noted that the number of rows of the first principal component in step 203 is a one-dimensional vector with a column number of 1, and by converting it into a w x h two-dimensional matrix, it is a two-dimensional matrix after dimension reduction; In this embodiment, the original infrared thermograph in step 102 is a gray scale image with a size of 428x586, after four rounds of cropping, the cropping ratio is not more than 1 / 3, after removing the edge influence of the ultrasonic excitation source and the experimental device, the obtained image size is a gray scale image with a size of 180x256, that is, w is 180 and h is 256.

[0035] In this embodiment, in specific implementation, the value of the hyperparameter The value range of the hyperparameter is (0, 1). can take different values to make According to the similarity matrix The difference value of the fusion similarity matrix , the weight of different steepness is generated. Further preferably, The value of 0.1 can be obtained according to The difference value can obtain a large weight coefficient to suppress the influence of the heating and cooling process.

[0036] In this embodiment, in the specific implementation, for the impact damage sample corresponding to 20J energy, as shown in Figure 2 The ultrasonic C-scan image is shown in (a), and the damage recognition results obtained by the FastCD method (discrete spectrum decomposition damage segmentation model), the WIASDMD method (wavelet sparse dictionary damage segmentation model), and the method of the present application are shown in (c)-(d), respectively. The average iteration time of the FastCD method, the WIASDMD method, and the method of the present application is 14.35 seconds, 10.28 seconds, and 2.898 seconds, respectively. The ultrasonic C-scan image is used as a reference for quantitative comparison, and the segmentation accuracy is 88.31%, 94.78%, and 96.72%, respectively.

[0037] For the impact damage sample corresponding to 40J energy, as shown in Figure 3 The ultrasonic C-scan image is shown in (a), and the damage recognition results obtained by the FastCD method (discrete spectrum decomposition damage segmentation model), the WIASDMD method (wavelet sparse dictionary damage segmentation model), and the method of the present application are shown in (c)-(d), respectively. The average iteration time of the FastCD method, the WIASDMD method, and the method of the present application is 18.31 seconds, 25.63 seconds, and 7.623 seconds, respectively. The segmentation accuracy is 91.07%, 90.45%, and 97.58%, respectively.

[0038] In summary, the method of the present application has simple steps and reasonable design. Based on the pixel anchor point similarity matrix and the fusion similarity matrix of each infrared thermal image, and considering the adaptive weighting weight of each infrared thermal image, an ultrasonic infrared thermal imaging time series self-weighting fusion model is constructed to obtain the optimal fusion similarity matrix and realize damage region segmentation. No manual annotation information is required, and the time series graph weight is dynamically adjusted to realize unsupervised device damage detection.

[0039] The above is only a preferred embodiment of the present application, and does not limit the present application. Any simple modification, change, and equivalent structure change according to the technical essence of the present application to the above embodiment are still within the protection scope of the technical solution of the present application.

Claims

1. A device damage detection method based on ultrasonic infrared thermography time series self-weighting, characterized by, The method comprises the following steps: Step one, the specific process is as follows: Step 101, using an ultrasonic infrared thermal image detection system to collect infrared thermal images of the equipment to be tested, and obtaining an ultrasonic infrared thermal imaging sequence; Step 102, using a computer to take the tth original infrared thermal image from the ultrasonic infrared thermal imaging sequence, and cropping the edges of the tth original infrared thermal image to obtain the tth cropped infrared thermal image, wherein the value of t ranges from 1 to N, and N represents the length of the ultrasonic infrared thermal imaging sequence; Step 103, repeating step 102 multiple times to complete the cropping of the Nth original infrared thermal image, and obtaining the Nth cropped infrared thermal image; Step 104, the first cropped infrared thermal image to the Nth cropped infrared thermal image are denoted as N infrared thermal images. Step two, the specific process is as follows:

2. A method for detecting damage to equipment based on ultrasonic infrared thermography time series self-weighting according to claim 1, characterized in that: Step 201, combining the N infrared thermal images according to the channel to obtain a high-dimensional infrared thermal image of w x h x N; wherein w represents the width of the cropped infrared thermal image, and h represents the height of the cropped infrared thermal image; Step 202, converting the high-dimensional infrared thermal image of w x h x N into a two-dimensional matrix; wherein the number of rows of the two-dimensional matrix is w x h, and the number of columns of the two-dimensional matrix is N; Step 203, using a computer to perform dimension reduction processing on the two-dimensional matrix using principal component analysis, and taking the first principal component as the two-dimensional matrix after dimension reduction; Step 204, using a computer to convert the two-dimensional matrix after dimension reduction into a two-dimensional grayscale image; Step 205, using a computer to segment the two-dimensional grayscale image according to the set number of superpixel blocks I' using the entropy rate superpixel segmentation method, and obtaining I' superpixel blocks; wherein I' is a positive integer.

3. A method for detecting damage in equipment based on ultrasonic infrared thermography time series self-weighting according to claim 2, characterized in that: Step three, the specific process is as follows: Step four, the specific process is as follows: Step five, the specific process is as follows: Step six, the specific process is as follows: Step 601, using a computer to set the number of categories to 2, and using the spectral clustering method to cluster process the optimal fusion similarity matrix to obtain pixel points with a category label of 1 and pixel points with a category label of 2; Step 602, using a computer to obtain the pixel mean of the pixel points with a category label of 1 and the pixel mean of the pixel points with a category label of 2, and the pixel region corresponding to the category label with the larger pixel mean is the damage region.

4. A method for detecting damage to equipment based on ultrasonic infrared thermography time series self-weighting according to claim 3, characterized in that: ​ Step 301, using a computer to obtain the anchor point of the i-th superpixel block on the t-th infrared thermal image according to ; wherein, ; wherein, represents the pixel value of the e-th pixel point in the i-th superpixel block on the t-th infrared thermal image, and the value range of e is , the value range of i is 1 to I', represents the total number of pixel points in the i-th superpixel block. Step 302, using a computer to calculate the similarity value between the jth pixel point on the tth infrared thermal image and the anchor point of the ith superpixel block according to ; wherein, the tth infrared thermal image is the tth infrared thermal image in the sequence of infrared thermal images; the jth pixel point is the jth pixel point in the tth infrared thermal image; the ith superpixel block is the ith superpixel block in the sequence of superpixel blocks; and the anchor point is the anchor point of the ith superpixel block. ; wherein, represents the pixel value of the jth pixel point on the tth infrared thermal image, and j ranges from 1 to w x h. Step 303, repeating step 302 multiple times to obtain the similarity value between each pixel point and each anchor point on the tth infrared thermal image, and combining the similarity value between each pixel point and each anchor point into a pixel-anchor similarity matrix of the tth infrared thermal image .

5. A method for detecting damage in equipment based on ultrasonic infrared thermography time series self-weighting according to claim 4, characterized in that: ​ The computer is used to set N infrared thermal images corresponding to the fusion similarity matrix as , input the pixel anchor point similarity matrix of the tth infrared thermal image , and construct an ultrasonic infrared thermal imaging time sequence self-weighted fusion model ; wherein min represents the minimum value, s.t. represents the constraint condition, represents the square of the F norm, represents the hyperparameter, represents the natural logarithm function, has w x h rows and I' columns; is a first constant matrix, and each element in the matrix is 1, and has I' x 1 size; is a second constant matrix, and each element in the matrix is 1, and has w x h rows and 1 column; represents the adaptive weighting weight during fusion of the tth infrared thermal image.

6. A method for detecting damage in equipment based on ultrasonic infrared thermography time series self-weighting according to claim 5, characterized in that: ​ Step 501, adopt computer to make each infrared thermal image fusion when the initial value of adaptive weighting weight is ; Step 502, input the pixel anchor point similarity matrix of the tth infrared thermal image Let Substitute the ultrasonic infrared thermal imaging time sequence self-weighted fusion model to obtain the converted ultrasonic infrared thermal imaging time sequence self-weighted fusion model ; Step 503, solving the converted ultrasonic infrared thermal imaging time sequence self-weighted fusion model to obtain an optimal fusion similarity matrix after first iteration ; Step 504, using a computer to obtain the adaptive weighting weight in the fusion of the tth infrared thermal image after the first iteration according to ;​ Step 505, according to the method of step 502 to step 504, let , the adaptive weighting weight of the tth infrared thermal image fusion at the first iteration , the optimal fusion similarity matrix after the second iteration is obtained by substituting and solving , and the adaptive weighting weight of the tth infrared thermal image fusion at the second iteration ; Step 506, repeating step 505 multiple times, let The adaptive weighting weight of the tth infrared thermal image fusion at the k-1th iteration is substituted and solved to obtain the optimal fusion similarity matrix after the kth iteration and the adaptive weighting weight of the tth infrared thermal image fusion at the kth iteration ; wherein k is a positive integer greater than 2; Step 507, the difference between the minimum value of the objective function of the kth iteration and the minimum value of the objective function of the k-1th iteration is less than 0.01, then the iteration ends, and the optimal fusion similarity matrix after the kth iteration is obtained is denoted as the optimal fusion similarity matrix.

7. A method for detecting damage in equipment based on ultrasonic infrared thermography time series self-weighting according to claim 6, characterized in that: ​ ​ ​

Images