Surfacing weld defect detection method and device, electronic equipment and storage medium

Abstract

The invention discloses a surfacing weld defect detection method and device, electronic equipment and a storage medium, the method and device are applied to the electronic equipment and used for detecting defects of surfacing welds on a blade wear-resistant layer, and specifically, a training data set is constructed by collecting selectable surfacing weld images on the blade wear-resistant layer; the optional training data set comprises a training set and a test set; constructing a support vector machine model; respectively training and testing the selectable support vector machine model by using the selectable training data set to obtain a defect detection model; and inputting the surfacing weld image of the wear-resistant layer of the to-be-detected blade into the optional defect detection model for processing to obtain a defect detection result. According to the scheme, professionals do not need to determine whether defects exist or not and the types, positions, sizes and the like of the defects according to experience, so that the influence of various subjective and objective factors can be avoided, and the defect identification accuracy can be improved.

Description

Technical Field

[0001] The present application relates to the field of welding technology, and more specifically, to a method, device, electronic device and storage medium for detecting defects in cladding welds. Background Art

[0002] The quality of the weld seam on the blade's wear layer is crucial to its service life and can even significantly impact the safe operation of the engine or turbine in which it is used. Therefore, effective detection of weld defects is essential. In current production, the primary approach is to manually analyze images of the weld seam on the blade's wear layer, empirically determining the presence of defects and their type, location, and size. Even experienced professionals are susceptible to various subjective factors, such as visual fatigue and emotional changes from prolonged viewing of similar images, which can reduce the accuracy of defect identification. Summary of the Invention

[0003] In view of this, the present application provides a method, device, electronic device and storage medium for detecting defects in a build-up weld, which are used to detect the build-up weld on the wear-resistant layer of a blade based on a machine vision method to improve the accuracy of detection.

[0004] In order to achieve the above objectives, the following solutions are proposed:

[0005] A method for detecting defects in a cladding weld seam is applied to electronic equipment and is used to detect defects in a cladding weld seam on a blade wear-resistant layer. The method comprises the following steps:

[0006] A training data set is constructed by collecting images of the surfacing weld on the wear-resistant layer of the blade, wherein the training data set includes a training set and a test set;

[0007] Build a support vector machine model;

[0008] Using the training data set to train and test the support vector machine model, respectively, to obtain a defect detection model;

[0009] The surfacing weld image of the wear-resistant layer of the blade to be tested is input into the defect detection model for processing to obtain a defect detection result.

[0010] Optionally, the step of collecting images of the surfacing welds on the wear-resistant layer of the blade to construct a training data set comprises the following steps:

[0011] Collecting a plurality of the surfacing weld images to obtain an image data set including the plurality of the surfacing weld images;

[0012] Preprocessing each of the surfacing weld images to obtain a defect area;

[0013] Performing feature extraction on the pre-processed surfacing weld image to obtain feature parameters of the defect area;

[0014] The plurality of surfacing weld images are grouped according to a preset ratio to obtain the training set and the test set.

[0015] Optionally, the preprocessing of each of the surfacing weld images to obtain the defective area comprises the steps of:

[0016] Performing edge detection on each of the surfacing weld images;

[0017] performing image enhancement processing on the surfacing weld image based on an adaptive histogram;

[0018] Grayscale processing is performed on the weld grayscale image that has undergone image enhancement processing to obtain a weld grayscale image;

[0019] The weld grayscale image is binarized and segmented based on an adaptive threshold value to obtain a plurality of defect areas.

[0020] Optionally, the characteristic parameters include part or all of the defect aspect ratio, defect absolute grayscale mean difference, defect sharpness, defect absolute distance and defect circularity.

[0021] Optionally, the support vector model includes four layers of classification, namely, a first layer of classification, a second layer of classification, a third layer of classification, and a fourth layer of classification, wherein:

[0022] The first level of classification is used to distinguish between strip defects and circular defects based on the defect aspect ratio;

[0023] The second level classification is used to distinguish circular defects into circular slag defects and porosity defects based on the mean absolute grayscale difference of the defects, and to distinguish strip defects into crack defects and non-crack defects based on the sharpness of the defects;

[0024] The third level classification is used to distinguish non-crack defects into incomplete penetration defects and non-incomplete penetration defects according to the absolute distance of the defects;

[0025] The fourth level of classification is used to distinguish non-incomplete penetration defects into incomplete fusion defects and strip-shaped slag defects according to defect circularity.

[0026] Optionally, the method of using the training data set to train and test the support vector machine model to obtain a defect detection model includes the following steps:

[0027] Marking a sample defect location and a sample defect type for each data sample in the training data set;

[0028] Inputting the training set into the support vector machine model for training to obtain a defect detection model to be tested;

[0029] The defect detection model to be tested is tested and adjusted using the test set to obtain the defect detection model.

[0030] Optionally, the method of using the test set to test and adjust the defect detection model to be tested to obtain the defect detection model includes the following steps:

[0031] Inputting the test set into the defect detection model to be tested to perform defect identification, and obtain the identified defect location and the identified defect type;

[0032] Calculating an identification loss, wherein the identification loss includes a first identification loss between the sample defect position and the identified defect position, and a second identification loss between the sample defect type and the identified defect type;

[0033] The recognition loss is judged. If the recognition loss can meet the accuracy requirement, the test is terminated, and the defect detection model to be tested is output as the defect detection model;

[0034] If the recognition loss cannot meet the accuracy requirement, the model parameters of the defect detection model to be tested are adjusted, and the process returns to the step of inputting the test set into the defect detection model to be tested for defect recognition, obtaining the steps of identifying the defect location and identifying the defect type, and performing iterative training.

[0035] A surfacing weld defect detection device is applied to electronic equipment and is used to detect defects in surfacing welds on a blade wear-resistant layer. The surfacing weld defect detection device includes:

[0036] A data set construction module is configured to construct a training data set by collecting images of the surfacing weld on the wear-resistant layer of the blade, wherein the training data set includes a training set and a test set;

[0037] A model building module, configured to build a support vector machine model;

[0038] A model training module is configured to train and test the support vector machine model using the training data set to obtain a defect detection model;

[0039] The detection execution module is configured to input the surfacing weld image of the wear-resistant layer of the blade to be tested into the defect detection model for processing to obtain a defect detection result.

[0040] An electronic device comprising at least one processor and a memory connected to the processor, wherein:

[0041] The memory is used to store computer programs or instructions;

[0042] The processor is used to execute the computer program or instruction so that the electronic device can implement the above-mentioned method for detecting defects in surfacing welds.

[0043] A storage medium is applied to an electronic device, wherein the storage medium carries one or more computer programs, and the one or more computer programs can be executed by the electronic device, thereby enabling the electronic device to implement the above-mentioned method for detecting defects in surfacing welds.

[0044] As can be seen from the above technical solutions, the present application discloses a method, device, electronic device and storage medium for detecting defects in overlay welds. The method and device are applied to electronic devices and are used to detect defects in overlay welds on the wear-resistant layer of blades. Specifically, a training data set is constructed by collecting optional overlay weld images on the wear-resistant layer of blades. The optional training data set includes a training set and a test set; a support vector machine model is constructed; the optional support vector machine model is trained and tested using the optional training data set to obtain a defect detection model; the overlay weld image of the wear-resistant layer of the blade to be tested is input into the optional defect detection model for processing to obtain a defect detection result. This solution does not require professionals to determine the presence or absence of defects and the type, location, size, etc. of the defects based on experience. Therefore, it can avoid being affected by various subjective and objective factors, thereby improving the accuracy of defect identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0046] Figure 1 This is a flow chart of a method for detecting defects in a cladding weld according to an embodiment of the present application;

[0047] Figure 2 Flowchart of a method for constructing a training data set in this embodiment;

[0048] Figure 3 Flowchart of the training process of the defect detection model according to an embodiment of the present application;

[0049] Figure 4 This is a block diagram of a device for detecting defects in a cladding weld according to an embodiment of the present application;

[0050] Figure 5 This is a block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0052] The present application provides a solution for detecting the weld quality of the surfacing weld on the wear-resistant layer of the blade. The method obtains a defect detection model based on machine learning, and detects the surfacing weld of the wear-resistant layer of the blade to be tested based on the defect detection model to obtain the detection results. The specific solution is as follows.

[0053] Figure 1 This is a flow chart of a method for detecting defects in a cladding weld according to an embodiment of the present application.

[0054] like Figure 1 As shown, the cladding weld defect detection method provided in this embodiment is applied to an electronic device for detecting defects in the cladding weld based on a machine vision method. The electronic device can be understood as a computer, server, or cloud platform with data computing and information processing capabilities. The cladding weld defect detection method includes the following steps:

[0055] S1. Construct a training dataset based on the collected cladding weld images.

[0056] That is, by collecting multiple images of surfacing welds on the wear-resistant layer of blades, a training data set for training defect detection models is constructed. The training data set here includes a training set and a test set. The training set is used for preliminary training of the model, and the test set is used to test the preliminarily trained model to ensure that the processing accuracy of the model meets the requirements. Figure 2 As shown in Figure 2, the specific process of constructing the training dataset is as follows.

[0057] S101. Capture multiple cladding weld images using an industrial camera.

[0058] The collected surfacing weld images are constructed into an image dataset. The acquisition tool here is not limited to industrial cameras, but can also be other image acquisition devices, such as ordinary cameras or webcams.

[0059] S102: Preprocess each surfacing weld image.

[0060] By preprocessing each cladding weld image, at least one or more defect areas are obtained on each cladding weld image. Of course, there may be relatively perfect cladding weld images without defect areas. The specific process is as follows:

[0061] First, edge detection is performed on each surfacing weld image to determine the weld edge and the edge where defects occur in the image.

[0062] Then, the surfacing weld image is enhanced based on the adaptive histogram to improve the clarity, sharpness and contrast of the image, which is beneficial to the subsequent further processing.

[0063] Then, the weld grayscale image that has undergone image enhancement processing is grayscale processed to obtain a weld grayscale image;

[0064] Finally, the grayscale image of the weld overlay seam is binarized and segmented based on the adaptive threshold value, thereby finally obtaining one or more defect areas of each weld overlay seam image.

[0065] S103: Extract features from the pre-processed cladding weld image.

[0066] The purpose of feature extraction is to obtain the image parameters of the defect area of ​​the surfacing weld image. The feature parameters include but are not limited to part or all of the defect aspect ratio, defect absolute grayscale mean difference, defect sharpness, defect absolute distance and defect circularity.

[0067] S104: Group the surfacing weld images to obtain a training set and a test set.

[0068] After the above processing, the cladding weld images in the training dataset containing the above-mentioned feature parameters are grouped according to a certain ratio, that is, the training dataset is divided into a training set and a test set. For example, 80% of the cladding weld images in the training dataset can be constructed as the training set, and the remaining 20% ​​of the cladding weld images can be constructed as the test set for testing the defect detection model.

[0069] S2. Build a support vector machine model.

[0070] That is, a support vector machine model based on machine vision is constructed to monitor the defects of surfacing welds on the wear-resistant layer of blades. The model introduces a multi-scale spatial enhanced attention mechanism to adaptively assign appropriate weights to feature parameters of different scales.

[0071] The support vector machine model includes four classification layers. The first layer distinguishes strip defects from circular defects based on the defect aspect ratio. The second layer distinguishes circular defects into circular slag defects and porosity defects based on the defect absolute grayscale mean difference, and distinguishes strip defects into crack defects and non-crack defects based on the defect sharpness. The third layer distinguishes non-crack defects into incomplete penetration defects and non-incomplete penetration defects based on the defect absolute distance. The fourth layer distinguishes non-incomplete penetration defects into incomplete fusion defects and strip slag defects based on the defect circularity.

[0072] S3. Use the training dataset to train and test the support vector machine model.

[0073] By training and testing the support vector machine model, a defect detection model for defect detection of the weld seam image to be tested is obtained. The specific training process is as follows: Figure 3 shown.

[0074] S301: Label each data sample in the training data set.

[0075] That is, each cladding weld image in the training dataset is labeled, the defect area in each image is marked, and the sample defect location and sample defect type are marked. The labeling here includes not only labeling the cladding weld images in the training set, but also labeling each cladding weld image in the test set.

[0076] S302: Train the support vector machine model based on the training set.

[0077] That is, the training set is input into the support vector machine model for training processing, so that the undetermined parameters in the model can be calibrated through training, thereby obtaining the defect detection model to be tested. Specifically:

[0078] The defect aspect ratio is input into the support vector machine model for first-level classification. When the defect aspect ratio is greater than or equal to the preset aspect ratio threshold, the defect area is defined as a strip defect. When the defect aspect ratio is less than the preset aspect ratio threshold, the defect area is defined as a circular defect.

[0079] The absolute grayscale mean difference of circular defects is input into the support vector machine model for the second-level circular classification. When the absolute grayscale mean difference of the defect is negative, the circular defect is defined as a circular slag defect; when the absolute grayscale mean difference of the defect is positive, the circular defect is defined as a porosity defect.

[0080] The defect sharpness of the strip defect is input into the support vector machine model for second-level strip classification. When the defect sharpness is greater than or equal to a preset sharpness threshold, the strip defect is defined as a crack defect. When the defect sharpness is less than the preset sharpness threshold, the strip defect is defined as a non-crack defect.

[0081] The absolute distance of the non-crack defect is input into the support vector machine model for third-level classification. When the absolute distance of the defect is less than the preset absolute distance threshold, the non-crack defect is defined as an incomplete penetration defect. When the absolute distance of the defect is greater than or equal to the preset absolute distance threshold, the non-crack defect is defined as a non-incomplete penetration defect.

[0082] The defect circularity of non-lack of penetration defects is input into the support vector set model for the fourth-level classification. When the defect circularity is greater than or equal to the preset circularity threshold, the non-lack of penetration defect is defined as an incomplete fusion defect. When the defect circularity is less than the preset circularity threshold, the non-lack of penetration defect is defined as a strip-shaped slag defect.

[0083] S303: Use the test set to test and adjust the defect detection model to be tested.

[0084] That is, after completing the above preliminary training, the above-mentioned defect detection model to be tested is tested using the test set, and the parameters are adjusted according to the test results. Then, the test and adjustment are performed again. After multiple rounds of iteration, a defect detection model that meets the requirements is finally obtained. The specific process is as follows:

[0085] First, the labeled surfacing weld images in the test set are input into the defect detection model to be tested for defect recognition, and the identified defect locations and defect types are obtained.

[0086] Then, the recognition loss is calculated based on the recognized defect position and the sample defect position, and the recognized defect type and the sample defect type. The recognition loss includes a first recognition loss between the sample defect position and the recognized defect position, and a second recognition loss between the sample defect type and the recognized defect type.

[0087] Next, the recognition loss is determined. If it meets the preset accuracy requirements, the test ends and the defect detection model to be tested is output as a defect detection model. Accuracy requirements include, but are not limited to, the accuracy of type recognition, as well as the positioning and area accuracy of defect locations.

[0088] Finally, if the above judgment determines that the recognition loss cannot meet the accuracy requirement, the model parameters of the defect detection model to be tested are adjusted, and the test set is input into the defect detection model to be tested again for defect recognition, and the recognition loss is calculated again for re-judgment.

[0089] S4. Defect recognition is performed on the surfacing weld image of the wear-resistant layer of the blade to be tested.

[0090] That is, on the basis of obtaining the above-mentioned defect recognition model, when a user defect recognition request is received, the image of the surfacing weld of the wear-resistant layer of the blade to be tested input by the user is input into the defect recognition model, so that the model detects the defects and the user obtains the defect detection results output by it, which at least include the defect location and defect type.

[0091] As can be seen from the above technical solution, this embodiment provides a method for detecting defects in overlay welds, which is applied to electronic equipment and is used to detect defects in overlay welds on the wear-resistant layer of a blade. Specifically, a training data set is constructed by collecting optional overlay weld images on the wear-resistant layer of the blade, and the optional training data set includes a training set and a test set; a support vector machine model is constructed; the optional support vector machine model is trained and tested using the optional training data set to obtain a defect detection model; and the overlay weld image of the wear-resistant layer of the blade to be tested is input into the optional defect detection model for processing to obtain a defect detection result. This solution does not require professionals to determine the presence or absence of defects and the type, location, size, etc. of the defects based on experience. Therefore, it can avoid being affected by various subjective and objective factors, thereby improving the accuracy of defect identification.

[0092] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the module, program segment, or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order than that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of the boxes in the block diagram and / or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0093] Although the operations are depicted in a particular order, this should not be understood as requiring that the operations be performed in the particular order shown or in a sequential order.Multitasking and parallel processing may be advantageous under certain circumstances.

[0094] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0095] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages, or a combination thereof, including, but not limited to, object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as C or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer.

[0096] Figure 4 This is a block diagram of a cladding weld defect detection device according to an embodiment of the present application.

[0097] like Figure 4 As shown, the cladding weld defect detection device provided in this embodiment is applied to an electronic device for detecting defects in cladding welds based on a machine vision method. The electronic device can be understood as a computer, server or cloud platform with data computing and information processing capabilities. The cladding weld defect detection device specifically includes a data set construction module 10, a model construction module 20, a model training module 30 and a detection execution module.

[0098] The dataset construction module is used to construct a training dataset based on the collected cladding weld images.

[0099] That is, by collecting multiple images of surfacing welds on the wear-resistant layer of blades, a training data set for training defect detection models is constructed. The training data set here includes a training set and a test set. The training set is used for preliminary training of the model, and the test set is used to test the preliminarily trained model to ensure that the processing accuracy of the model meets the requirements. Figure 2 As shown in Figure 2, the specific process of constructing the training dataset is as follows.

[0100] S101. Capture multiple cladding weld images using an industrial camera.

[0101] The collected cladding weld images are constructed into an image dataset.

[0102] S102: Preprocess each surfacing weld image.

[0103] By preprocessing each cladding weld image, at least one or more defect areas are obtained on each cladding weld image. Of course, there may be relatively perfect cladding weld images without defect areas. The specific process is as follows:

[0104] First, edge detection is performed on each surfacing weld image to determine the weld edge and the edge where defects occur in the image.

[0105] Then, the surfacing weld image is enhanced based on the adaptive histogram to improve the clarity, sharpness and contrast of the image, which is beneficial to the subsequent further processing.

[0106] Then, the weld grayscale image that has undergone image enhancement processing is grayscale processed to obtain a weld grayscale image;

[0107] Finally, the grayscale image of the weld overlay seam is binarized and segmented based on the adaptive threshold value, thereby finally obtaining one or more defect areas of each weld overlay seam image.

[0108] S103: Extract features from the pre-processed cladding weld image.

[0109] The purpose of feature extraction is to obtain the image parameters of the defect area of ​​the surfacing weld image. The feature parameters include but are not limited to part or all of the defect aspect ratio, defect absolute grayscale mean difference, defect sharpness, defect absolute distance and defect circularity.

[0110] S104: Group the surfacing weld images to obtain a training set and a test set.

[0111] After the above processing, the cladding weld images in the training dataset containing the above-mentioned feature parameters are grouped according to a certain ratio, that is, the training dataset is divided into a training set and a test set. For example, 80% of the cladding weld images in the training dataset can be constructed as the training set, and the remaining 20% ​​of the cladding weld images can be constructed as the test set for testing the defect detection model.

[0112] The model building module is used to build the support vector machine model.

[0113] That is, a support vector machine model based on machine vision is constructed to monitor the defects of surfacing welds on the wear-resistant layer of blades. The model introduces a multi-scale spatial enhanced attention mechanism to adaptively assign appropriate weights to feature parameters of different scales.

[0114] The support vector machine model includes four classification layers. The first layer distinguishes strip defects from circular defects based on the defect aspect ratio. The second layer distinguishes circular defects into circular slag defects and porosity defects based on the defect absolute grayscale mean difference, and distinguishes strip defects into crack defects and non-crack defects based on the defect sharpness. The third layer distinguishes non-crack defects into incomplete penetration defects and non-incomplete penetration defects based on the defect absolute distance. The fourth layer distinguishes non-incomplete penetration defects into incomplete fusion defects and strip slag defects based on the defect circularity.

[0115] The model training module is used to train and test the support vector machine model using the training dataset.

[0116] By training and testing the support vector machine model, a defect detection model for defect detection of the weld seam image to be tested is obtained. The specific training process is as follows: Figure 3 shown.

[0117] S301: Label each data sample in the training data set.

[0118] That is, each cladding weld image in the training dataset is labeled, the defect area in each image is marked, and the sample defect location and sample defect type are marked. The labeling here includes not only labeling the cladding weld images in the training set, but also labeling each cladding weld image in the test set.

[0119] S302: Train the support vector machine model based on the training set.

[0120] That is, the training set is input into the support vector machine model for training processing, so that the undetermined parameters in the model can be calibrated through training, thereby obtaining the defect detection model to be tested. Specifically:

[0121] The defect aspect ratio is input into the support vector machine model for first-level classification. When the defect aspect ratio is greater than or equal to the preset aspect ratio threshold, the defect area is defined as a strip defect. When the defect aspect ratio is less than the preset aspect ratio threshold, the defect area is defined as a circular defect.

[0122] The absolute grayscale mean difference of circular defects is input into the support vector machine model for the second-level circular classification. When the absolute grayscale mean difference of the defect is negative, the circular defect is defined as a circular slag defect; when the absolute grayscale mean difference of the defect is positive, the circular defect is defined as a porosity defect.

[0123] The defect sharpness of the strip defect is input into the support vector machine model for second-level strip classification. When the defect sharpness is greater than or equal to a preset sharpness threshold, the strip defect is defined as a crack defect. When the defect sharpness is less than the preset sharpness threshold, the strip defect is defined as a non-crack defect.

[0124] The absolute distance of the non-crack defect is input into the support vector machine model for third-level classification. When the absolute distance of the defect is less than the preset absolute distance threshold, the non-crack defect is defined as an incomplete penetration defect. When the absolute distance of the defect is greater than or equal to the preset absolute distance threshold, the non-crack defect is defined as a non-incomplete penetration defect.

[0125] The defect circularity of non-lack of penetration defects is input into the support vector set model for the fourth-level classification. When the defect circularity is greater than or equal to the preset circularity threshold, the non-lack of penetration defect is defined as an incomplete fusion defect. When the defect circularity is less than the preset circularity threshold, the non-lack of penetration defect is defined as a strip-shaped slag defect.

[0126] S303: Use the test set to test and adjust the defect detection model to be tested.

[0127] That is, after completing the above preliminary training, the above-mentioned defect detection model to be tested is tested using the test set, and the parameters are adjusted according to the test results. Then, the test and adjustment are performed again. After multiple rounds of iteration, a defect detection model that meets the requirements is finally obtained. The specific process is as follows:

[0128] First, the labeled surfacing weld images in the test set are input into the defect detection model to be tested for defect recognition, and the identified defect locations and defect types are obtained.

[0129] Then, the recognition loss is calculated based on the recognized defect position and the sample defect position, and the recognized defect type and the sample defect type. The recognition loss includes a first recognition loss between the sample defect position and the recognized defect position, and a second recognition loss between the sample defect type and the recognized defect type.

[0130] Next, the recognition loss is determined. If it meets the preset accuracy requirements, the test ends and the defect detection model to be tested is output as a defect detection model. Accuracy requirements include, but are not limited to, the accuracy of type recognition, as well as the positioning and area accuracy of defect locations.

[0131] Finally, if the above judgment determines that the recognition loss cannot meet the accuracy requirement, the model parameters of the defect detection model to be tested are adjusted, and the test set is input into the defect detection model to be tested again for defect recognition, and the recognition loss is calculated again for re-judgment.

[0132] The detection execution module is used to identify defects in the surfacing weld image of the wear-resistant layer of the blade to be tested.

[0133] That is, on the basis of obtaining the above-mentioned defect recognition model, when a user defect recognition request is received, the image of the surfacing weld of the wear-resistant layer of the blade to be tested input by the user is input into the defect recognition model, so that the model detects the defects and the user obtains the defect detection results output by it, which at least include the defect location and defect type.

[0134] As can be seen from the above technical solution, this embodiment provides a device for detecting defects in overlay welds, which is applied to electronic equipment and is used to detect defects in overlay welds on the wear-resistant layer of a blade. Specifically, a training data set is constructed by collecting optional overlay weld images on the wear-resistant layer of the blade, and the optional training data set includes a training set and a test set; a support vector machine model is constructed; the optional support vector machine model is trained and tested using the optional training data set to obtain a defect detection model; and the overlay weld image of the wear-resistant layer of the blade to be tested is input into the optional defect detection model for processing to obtain a defect detection result. This solution does not require professionals to determine the presence or absence of defects and the type, location, size, etc. of the defects based on experience. Therefore, it can avoid being affected by various subjective and objective factors, thereby improving the accuracy of defect identification.

[0135] The units involved in the embodiments described in this disclosure may be implemented in software or hardware. In some cases, the name of a unit does not limit the unit itself. For example, the first acquisition unit may also be described as a "unit for acquiring at least two Internet Protocol addresses."

[0136] The functions described above herein may be performed, at least in part, by one or more hardware logic components. For example, and without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chip (SOCs), complex programmable logic devices (CPLDs), and the like.

[0137] Figure 5 This is a block diagram of an electronic device according to an embodiment of the present application.

[0138] Reference below Figure 5 , which shows a schematic diagram of the structure of an electronic device suitable for implementing the embodiments of the present disclosure. The terminal device in the embodiments of the present disclosure may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (such as in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. This electronic device is merely an example and should not limit the functionality and scope of use of the embodiments of the present disclosure.

[0139] The electronic device may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory ROM 502 or a program loaded from an input device 506 into a random access memory RAM 503. Various programs and data required for the operation of the electronic device are also stored in the RAM. The processing device, ROM, and RAM are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0140] Typically, the following devices may be connected to the I / O interface: input devices such as a touch screen, touchpad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 507 such as a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 508 such as a magnetic tape, hard disk, etc.; and communication devices 509. Communication devices 509 may allow the electronic device to communicate with other devices wirelessly or by wire to exchange data. Although the figures illustrate electronic devices with various devices, it should be understood that not all of the devices shown are required to be implemented or present. More or fewer devices may be implemented or present instead.

[0141] The present application also provides a computer-readable storage medium embodiment.

[0142] The computer-readable storage medium is applied to an electronic device and carries one or more computer programs. When the one or more computer programs are executed by the electronic device, the electronic device is used to detect defects in the surfacing welds on the wear-resistant layer of the blade. Specifically, a training data set is constructed by collecting optional images of the surfacing welds on the wear-resistant layer of the blade. The optional training data set includes a training set and a test set; a support vector machine model is constructed; the optional support vector machine model is trained and tested using the optional training data set to obtain a defect detection model; and the surfacing weld images of the wear-resistant layer of the blade to be tested are input into the optional defect detection model for processing to obtain defect detection results. This solution does not require professionals to determine the presence or absence of defects and the type, location, size, etc. of the defects based on experience. Therefore, it can avoid being affected by various subjective and objective factors, thereby improving the accuracy of defect identification.

[0143] It should be noted that the computer-readable medium disclosed herein may be a computer-readable signal medium or a computer-readable storage medium or any combination thereof. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0144] In the present disclosure, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the foregoing.

[0145] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0146] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they become aware of the basic creative concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0147] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or terminal device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or terminal device that includes the element.

[0148] The technical solution provided by the present invention is introduced in detail above. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for detecting defects in a cladding weld seam, applied to electronic equipment, for detecting defects in a cladding weld seam on a blade wear-resistant layer, characterized in that: The method for detecting defects in surfacing welds comprises the following steps: A training data set is constructed by collecting images of the surfacing weld on the wear-resistant layer of the blade, wherein the training data set includes a training set and a test set; Build a support vector machine model; Using the training data set to train and test the support vector machine model, respectively, to obtain a defect detection model; The surfacing weld image of the wear-resistant layer of the blade to be tested is input into the defect detection model for processing to obtain a defect detection result.

2. The method for detecting defects in a cladding weld according to claim 1, wherein: The method of collecting images of the surfacing weld seams on the blade wear-resistant layer to construct a training data set comprises the following steps: Collecting a plurality of the surfacing weld images to obtain an image data set including the plurality of the surfacing weld images; Preprocessing each of the surfacing weld images to obtain a defect area; Performing feature extraction on the pre-processed surfacing weld image to obtain feature parameters of the defect area; The plurality of surfacing weld images are grouped according to a preset ratio to obtain the training set and the test set.

3. The method for detecting defects in a cladding weld according to claim 2, wherein: The method of pre-processing each of the surfacing weld images to obtain the defect area comprises the following steps: Performing edge detection on each of the surfacing weld images; performing image enhancement processing on the surfacing weld image based on an adaptive histogram; Grayscale processing is performed on the weld grayscale image that has undergone image enhancement processing to obtain a weld grayscale image; The weld grayscale image is binarized and segmented based on an adaptive threshold value to obtain a plurality of defect areas.

4. The method for detecting the cladding weld area according to claim 2, wherein: The characteristic parameters include part or all of the defect aspect ratio, defect absolute grayscale mean difference, defect sharpness, defect absolute distance and defect circularity.

5. The method for detecting defects in a cladding weld according to claim 1, wherein: The support vector model includes four layers of classification, namely, first layer classification, second layer classification, third layer classification and fourth layer classification, wherein: The first level of classification is used to distinguish between strip defects and circular defects based on the defect aspect ratio; The second level classification is used to distinguish circular defects into circular slag defects and porosity defects based on the mean absolute grayscale difference of the defects, and to distinguish strip defects into crack defects and non-crack defects based on the sharpness of the defects; The third level classification is used to distinguish non-crack defects into incomplete penetration defects and non-incomplete penetration defects according to the absolute distance of the defects; The fourth level of classification is used to distinguish non-incomplete penetration defects into incomplete fusion defects and strip-shaped slag defects according to defect circularity.

6. The method for detecting defects in a cladding weld according to claim 1, wherein: The method of using the training data set to train and test the support vector machine model to obtain a defect detection model includes the following steps: Marking a sample defect location and a sample defect type for each data sample in the training data set; Inputting the training set into the support vector machine model for training to obtain a defect detection model to be tested; The defect detection model to be tested is tested and adjusted using the test set to obtain the defect detection model.

7. The method for detecting defects in a cladding weld according to claim 6, wherein: The method of using the test set to test and adjust the defect detection model to be tested to obtain the defect detection model includes the following steps: Inputting the test set into the defect detection model to be tested to perform defect identification, and obtain the identified defect location and the identified defect type; Calculating an identification loss, wherein the identification loss includes a first identification loss between the sample defect position and the identified defect position, and a second identification loss between the sample defect type and the identified defect type; The recognition loss is judged. If the recognition loss can meet the accuracy requirement, the test is terminated, and the defect detection model to be tested is output as the defect detection model; If the recognition loss cannot meet the accuracy requirement, the model parameters of the defect detection model to be tested are adjusted, and the process returns to the step of inputting the test set into the defect detection model to be tested for defect recognition, obtaining the steps of identifying the defect location and identifying the defect type, and performing iterative training.

8. A cladding weld defect detection device, applied to electronic equipment, for detecting defects in cladding welds on a blade wear-resistant layer, characterized in that: The surfacing weld defect detection device comprises: A data set construction module is configured to construct a training data set by collecting images of the surfacing weld on the wear-resistant layer of the blade, wherein the training data set includes a training set and a test set; A model building module, configured to build a support vector machine model; A model training module is configured to train and test the support vector machine model using the training data set to obtain a defect detection model; The detection execution module is configured to input the surfacing weld image of the wear-resistant layer of the blade to be tested into the defect detection model for processing to obtain a defect detection result.

9. An electronic device, characterized in that: The electronic device comprises at least one processor and a memory connected to the processor, wherein: The memory is used to store computer programs or instructions; The processor is used to execute the computer program or instruction so that the electronic device implements the method for detecting defects in a cladding weld according to any one of claims 1 to 7.

10. A storage medium, applied to an electronic device, characterized in that: The storage medium carries one or more computer programs, and the one or more computer programs can be executed by the electronic device, so that the electronic device can implement the method for detecting defects in a cladding weld as described in any one of claims 1 to 7.

Images