A wind power generation blade defect intelligent comparison method and device and electronic equipment

Abstract

The application provides a wind power generation blade defect intelligent comparison method and device and electronic equipment, and relates to the technical field of defect detection. The method comprises the following steps: acquiring a blade defect video; frame-by-frame dividing the blade defect video to obtain a plurality of minimum granularity frame pictures; identifying defect feature elements in the minimum granularity frame pictures; marking the positions of the defect feature elements in the minimum granularity frame pictures; matching the defect feature elements in the minimum granularity frame pictures with a constructed blade defect feature element comparison library to obtain a defect type comparison result; and feeding back the defect type comparison result to a user end. The application realizes identification result and rapid feedback, reduces manual detection omission, effectively improves identification accuracy, and solves problems such as short identification time, short repair time, and difficult comparison of multi-layer paint history quality retention caused by layered construction.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of defect detection, in particular to a wind power blade defect intelligent comparison method and device and electronic equipment. BACKGROUND

[0002] The wind power blade protection generally adopts various modes such as film covering and protective paint, and the operation of the artificial operation in the protective paint construction process requires proficiency, speed and accuracy, and a plurality of people are required to be highly tacit in the construction, and the wind speed, humidity and atmospheric pressure of the processing environment also have certain requirements. The high standard requirement of the protective paint surface is to adapt to the diversity of the wind power blade in the complex natural environment, so as to maintain the economic benefit of the blade rotation power generation for a long time, and reduce the frequent operation and maintenance and inspection cost caused by external factors.

[0003] The wind power blade protective paint construction surface quality index detection includes a plurality of detection indexes such as smoothness, flatness, tower joint and damage of the protective paint construction surface, and at present, the unit particle size operation surface can only be identified by the naked eye of the construction personnel. Due to the sunlight refraction, light irradiation, identification angle and detection surface height and other external factors, there are problems of slow quality detection speed, high accuracy and high omission rate for different parts of the blade.

[0004] The wind power blade protective paint construction process is multi-pass brushing, and the quality detection of each layer of protective paint operation is the key and difficulty of quality guarantee. If the problem in the construction cannot be found and corrected in time, the problems existing in the previous construction surface will be covered layer by layer, and only the outermost construction surface detection is left, the multi-layer tracking type paint surface construction quality detection cannot be achieved, and the irreversible quality traceability detection problem exists at the same time.

[0005] The wind power blade is large in size and long in length, and various transportation and hoisting problems such as scratches, scratches, bruises and extrusion are prone to occur during transportation. The difficulty of the blade protective paint quality detection becomes more difficult with the change of the professional and environmental conditions.

[0006] Therefore, the present application provides a wind power blade defect intelligent comparison method, device and electronic equipment. SUMMARY

[0007] The present application provides a wind power blade defect intelligent comparison method, device and electronic equipment, realizes the identification result and fast feedback, reduces the artificial detection omission, effectively improves the identification accuracy, and solves the problems of short identification time, short repair time, multi-layer paint surface history quality comparison difficulty and the like caused by the layer-by-layer construction.

[0008] The present specification provides a file uploading method, comprising:

[0009] Obtaining a blade defect video;

[0010] Frame-by-frame dividing the blade defect video to obtain a plurality of minimum granularity frame pictures;

[0011] Identifying defect feature elements in the minimum granularity frame pictures;

[0012] Marking the positions of the defect feature elements in the minimum granularity frame pictures;

[0013] Matching the defect feature elements in the minimum granularity frame pictures with a constructed blade defect feature element comparison library to obtain a defect type comparison result;

[0014] Feeding back the defect type comparison result to a user end.

[0015] Optionally, the constructing of the blade defect feature element comparison library comprises:

[0016] Identifying defect feature elements of existing blade defect pictures one by one;

[0017] Marking the positions and types of the defect feature elements of the existing blade defect pictures;

[0018] Classifying the blade defect pictures according to the types of the defect feature elements of the blade defect pictures to obtain a blade defect feature element comparison library.

[0019] Optionally, before the identifying of the defect feature elements of the blade defect pictures one by one, the method further comprises:

[0020] Obtaining a plurality of blade defect pictures to be processed;

[0021] Judging whether there are duplicate pictures and / or invalid pictures in the plurality of blade defect pictures to be processed;

[0022] When there are duplicate pictures and / or invalid pictures in the plurality of blade defect pictures to be processed, marking the duplicate pictures and / or invalid pictures;

[0023] Removing the marked blade defect pictures to be processed to obtain initial blade defect pictures.

[0024] Optionally, the identifying of the defect feature elements of the obtained blade defect pictures and the marking of the positions and types of the defect feature elements of the blade defect pictures comprise:

[0025] Identifying the defect feature elements of the blade defect pictures by using a convolutional neural network pattern recognition model and a blade defect feature element comparison library;

[0026] marking a defect feature element and a type of the defect feature element of the blade defect picture;

[0027] identifying a position of the defect feature element of the blade defect picture in the blade defect picture by using a convolutional neural network digital recognition model.

[0028] Optionally, before the judging whether the plurality of blade defect pictures to be processed exist repeated pictures and / or invalid pictures, the method further comprises: performing image processing on the plurality of blade defect pictures to be processed in at least one of edge cutting, rotation, stretching and image enhancement.

[0029] Optionally, the matching the defect feature element of the minimum granularity frame picture with the constructed blade defect feature element comparison library to obtain a defect type comparison result comprises:

[0030] matching the defect feature element of the minimum granularity frame picture with the constructed blade defect feature element comparison library to obtain a similarity of the defect feature element of the minimum granularity frame picture and different defect types;

[0031] judging whether the similarity of the defect feature element of the minimum granularity frame picture and different defect types exceeds a threshold value;

[0032] when the similarity of the defect feature element of the minimum granularity frame picture and different defect types exceeds the threshold value, sorting all the defect types with the similarity exceeding the threshold value to obtain a defect type comparison result.

[0033] The present specification provides a wind power blade defect intelligent comparison device, comprising:

[0034] an acquisition module configured to acquire a blade defect video;

[0035] a frame dividing module configured to divide the blade defect video frame by frame to obtain a plurality of minimum granularity frame pictures;

[0036] an identification module configured to identify a defect feature element in the minimum granularity frame picture;

[0037] a marking module configured to mark a position of the defect feature element of the minimum granularity frame picture;

[0038] a comparison module configured to match the defect feature element of the minimum granularity frame picture with a constructed blade defect feature element comparison library to obtain a defect type comparison result;

[0039] a feedback module configured to feed back the defect type comparison result to a user end.

[0040] Optionally, the constructing the blade defect feature element comparison library comprises:

[0041] identifying defect features of the existing blade defect pictures one by one;

[0042] marking positions and types of the defect features of the existing blade defect pictures;

[0043] classifying the blade defect pictures according to the types of the defect features of the blade defect pictures, to obtain a blade defect feature comparison library.

[0044] Optionally, before the defect features of the blade defect pictures are identified one by one, the method further comprises:

[0045] obtaining a plurality of blade defect pictures to be processed;

[0046] judging whether the plurality of blade defect pictures to be processed contain duplicate pictures and / or invalid pictures;

[0047] when the plurality of blade defect pictures to be processed contain duplicate pictures and / or invalid pictures, marking the duplicate pictures and / or invalid pictures;

[0048] eliminating the marked blade defect pictures to be processed, to obtain existing blade defect pictures.

[0049] Optionally, the identifying defect features of the blade defect pictures, marking positions and types of the defect features of the blade defect pictures, comprises:

[0050] identifying defect features of the blade defect pictures by using a convolutional neural network pattern recognition model and a blade defect feature comparison library;

[0051] marking the defect features of the blade defect pictures and types thereof;

[0052] identifying positions of the defect features of the blade defect pictures in the blade defect pictures by using a convolutional neural network digital recognition model.

[0053] Optionally, before the judgment whether the plurality of blade defect pictures to be processed contain duplicate pictures and / or invalid pictures, the method further comprises: performing image processing on the plurality of blade defect pictures to be processed by at least one of edge cutting, rotation, stretching and image enhancement.

[0054] Optionally, the comparison module comprises:

[0055] a matching unit, configured to match the defect features of the minimum granularity frame picture with the blade defect feature comparison library constructed, to obtain similarity of the defect features of the minimum granularity frame picture and different defect types;

[0056] The judgment unit is used to determine whether the similarity between different defect types and the defect feature elements of the framed image at the smallest granularity exceeds a threshold.

[0057] The sorting unit is used to sort all the defect types with similarity exceeding the threshold when the similarity between different defect types and the defect feature elements of the frame image at the smallest granularity exceeds the threshold, thereby obtaining the defect type comparison result.

[0058] This specification also provides an electronic device, wherein the electronic device includes:

[0059] Processor; and,

[0060] A memory that stores computer-executable instructions, which, when executed, cause the processor to perform any of the methods described above.

[0061] This specification also provides a computer-readable storage medium that stores one or more programs that, when executed by a processor, implement any of the methods described above.

[0062] This specification describes a method for receiving user-uploaded videos of blade defects. The videos are then segmented frame by frame to obtain several smallest-granularity frame images. A convolutional neural network image recognition model, combined with a blade defect feature element comparison library, is used to identify the defect feature elements in each frame image. The defect feature elements are circled on the defective blade images. Based on the matching results between the defect feature elements of the smallest-granularity frame images and the constructed blade defect feature element comparison library, a defect type comparison result is obtained. This result includes the defect type and location. The defect type comparison result is then sent to the user's device, allowing the user to determine the defect type of the wind turbine blade corresponding to the video. The user can use a mobile app, thus overcoming geographical limitations and receiving wind turbine blade defect type information as long as there is a mobile signal; alternatively, it can be transmitted via Wi-Fi, enabling real-time reception of wind turbine blade defect types. This invention effectively solves the problems of easy omissions during manual inspection, difficulty in accurate identification, and the challenges of short identification and repair times and difficulty in comparing historical quality data for multiple layers of paint due to layered construction. Attached Figure Description

[0063] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0064] Figure 1A principle schematic diagram of a wind power generation blade defect intelligent comparison method provided by an embodiment of the present specification;

[0065] Figure 2 A principle schematic diagram of constructing a blade defect feature element comparison library in step S150 of a wind power generation blade defect intelligent comparison method provided by an embodiment of the present specification;

[0066] Figure 3 A structure schematic diagram of a wind power generation blade defect intelligent comparison device provided by an embodiment of the present specification;

[0067] Figure 4 A structure schematic diagram of an electronic device provided by an embodiment of the present specification;

[0068] Figure 5 A principle schematic diagram of a computer readable medium provided by an embodiment of the present specification. DETAILED DESCRIPTION

[0069] The following description is provided so as to enable any person skilled in the art to make or use the present application. The preferred embodiments described in the following description are only examples for implementing the present application. Other apparent variants can be derived by those skilled in the art without departing from the spirit and scope of the present application. The essential principles of the present application defined in the following description can be applied to other embodiments, variants, improvements, equivalents and other technical solutions without departing from the spirit and scope of the present application.

[0070] The following description will be made in conjunction with the accompanying drawings. Figures 1-5 Exemplary embodiments of the present application are described more fully hereinafter with reference to the accompanying drawings. However, the exemplary embodiments can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these exemplary embodiments are provided so that the present application will be thorough and complete, and will fully convey the inventive concept to those skilled in the art. Like reference numerals refer to like elements throughout the specification. Repetitive descriptions of like elements will be omitted for brevity.

[0071] In the premise of conforming to the technical concept of the present application, the features, structures, characteristics or other details described in a certain specific embodiment do not exclude that they can be combined in one or more other embodiments in a suitable manner.

[0072] In the description of the specific embodiments, the features, structures, characteristics or other details described by the present application are to enable those skilled in the art to fully understand the embodiments. However, it does not exclude that one or more of the specific features, structures, characteristics or other details can be practiced by those skilled in the art without the technical solution of the present application.

[0073] The flowchart shown in the accompanying drawings is only an exemplary illustration, and is not necessarily required to include all the contents and operations / steps, nor is it necessarily required to be executed in the order described. For example, some operations / steps can be further decomposed, and some operations / steps can be combined or partially combined, so the actual execution order can be changed according to actual conditions.

[0074] The block diagram shown in the accompanying drawings is only a functional entity, and does not necessarily correspond to a physically independent entity. That is, the functional entity can be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0075] The term "and / or" or "and / or" includes all combinations of one or more of the associated listed items.

[0076] Figure 1 A schematic diagram of the principle of a wind power blade defect intelligent comparison method provided in the embodiments of the present specification, which can include:

[0077] S110: Obtain a blade defect video.

[0078] In the detailed description of the present specification, the blade video can be obtained by shooting with a high-definition 4K mobile phone, a video camera, a wearable device, a monitoring device, etc. The blade video can also be obtained by shooting with a drone carrying a video camera, a monitoring device, etc. The drone runs at a constant speed to obtain the blade video. The video with a preliminarily judged blade defect is selected to obtain the blade defect video.

[0079] S120: Frame the blade defect video one by one to obtain a plurality of minimum granularity frame pictures;

[0080] S130: Identify the defect feature elements in the minimum granularity frame pictures.

[0081] S140: Mark the positions of the defect feature elements of the minimum granularity frame pictures.

[0082] In the detailed description of the present specification, the blade defect video is frame-processed, that is, one frame is saved as one picture, to obtain a plurality of minimum granularity frame pictures. The minimum granularity frame pictures are subjected to defect feature element identification to determine the frame pictures with blade defects, and the defect feature elements are circled on the frame pictures with blade defects to obtain the position information of the defect feature elements. The acquisition of the position information of the defect feature elements specifically includes that the wind power blade is sprayed with a prominent scale mark, and the color of the scale mark can be selected according to actual needs, and there is a large color difference with the bottom color of the wind power blade. After the defect feature elements are circled, the position information of the defect feature elements can be obtained by referring to the scale mark of the wind power blade.

[0083] S150: match the defect feature elements of the minimum granularity sub-frame picture with the constructed leaf defect feature element matching library to obtain a defect type matching result.

[0084] In the detailed description of the present specification, the matching result of the defect feature elements of the minimum granularity sub-frame picture with the constructed leaf defect feature element matching library is obtained, and the defect type matching result includes a defect type and a defect position.

[0085] Optionally, the step S150 further includes:

[0086] matching the defect feature elements of the minimum granularity sub-frame picture with the constructed leaf defect feature element matching library to obtain a similarity of the defect feature elements of the minimum granularity sub-frame picture with defect feature elements of different defect types;

[0087] judging whether the similarity of the defect feature elements of different defect types with the minimum granularity sub-frame picture exceeds a threshold value;

[0088] when the similarity of the defect feature elements of different defect types with the minimum granularity sub-frame picture exceeds the threshold value, ranking all the defect types with the similarity exceeding the threshold value to obtain a defect type matching result.

[0089] In the detailed description of the present specification, only the defect types with the similarity exceeding the threshold value are ranked, and the ranking can be arranged in descending order of the similarity, thereby improving the convenience of the user viewing the matching result.

[0090] Optionally, Figure 2 A principle diagram of constructing a leaf defect feature element matching library in a wind power generation leaf defect intelligent matching method provided by the present specification is shown in the following figure, and the method further includes:

[0091] S151: identify defect feature elements of the existing leaf defect picture one by one.

[0092] In the detailed description of the present specification, the initial leaf picture can be obtained by shooting through a high-definition 4K mobile phone, a camera, a video camera, a wearable device, a monitoring device, etc. The initial leaf picture can also be obtained by a drone carrying a camera, a video camera, a monitoring device, etc. The drone runs at a constant speed to obtain the initial leaf picture. The leaf picture supports 1920 / 4K resolution and supports JPG and PNG format pictures. After the initial leaf picture is screened by a person, the picture with a preliminary judgment of existing leaf defects is selected to obtain an existing leaf defect picture. After comparison with the defect-free leaf picture, the position of the defect feature elements of the existing leaf defect is circled.

[0093] Optionally, before the step S151, it further includes:

[0094] obtain a plurality of leaf defect pictures to be processed;

[0095] determine whether the plurality of leaf defect pictures to be processed contain duplicate pictures and / or invalid pictures.

[0096] In the detailed description of the present specification, after obtaining the leaf defect pictures to be processed, the leaf defect pictures to be processed are checked for duplicates, and it is determined whether the leaf defect pictures to be processed contain invalid pictures, including blank pictures and pictures without defect feature elements compared with the pictures of leaves without defects.

[0097] Optionally, before the determination of whether the plurality of leaf defect pictures to be processed contain duplicate pictures and / or invalid pictures, the plurality of leaf defect pictures to be processed are subjected to at least one of image processing, including edge cutting, rotation, stretching, and image enhancement.

[0098] In the detailed description of the present specification, before the determination of whether the plurality of leaf defect pictures to be processed contain duplicate pictures and / or invalid pictures, the leaf defect pictures to be processed are subjected to image enhancement processing, so as to avoid invalid pictures due to distortion, insufficient brightness, etc., and improve the utilization rate of data.

[0099] When the plurality of leaf defect pictures to be processed contain duplicate pictures and / or invalid pictures, the duplicate pictures and / or invalid pictures are marked;

[0100] The leaf defect pictures to be processed with the marks are removed, and existing leaf defect pictures are obtained.

[0101] In the detailed description of the present specification, after obtaining the leaf defect pictures to be processed, the leaf defect pictures to be processed are checked for duplicates, and the redundant duplicate pictures are marked; the invalid pictures are also marked. The leaf defect pictures to be processed with the marks are removed, and the remaining leaf defect pictures to be processed without the marks are existing leaf defect pictures. In the detailed description of the present specification, the leaf defect pictures to be processed are processed first, so as to avoid that the leaf defect feature element comparison library contains many useless pictures, occupies the space of the leaf defect feature element comparison library, wastes resources, and further affects the accuracy of the leaf defect feature element comparison library. Meanwhile, the leaf defect pictures can be manually screened before this step, and then screened in this step; or the leaf defect pictures can be manually screened after this step. The above double mode operation improves the accuracy of the leaf defect feature element comparison library.

[0102] S152: mark the positions and types of the defect feature elements of the initial leaf defect pictures.

[0103] In the specific embodiments of the present specification, the acquisition of the position information of the defect feature element specifically includes that the wind power blade is sprayed with a conspicuous scale mark, and the color of the scale mark can be selected according to actual needs, and there is a large color difference with the bottom color of the wind power blade. After the defect feature element is circled, the position information of the defect feature element is obtained by referring to the scale mark of the wind power blade.

[0104] According to the circled defect feature element, the defect feature element is classified according to the type of the defect feature element, and the type of the defect feature element mainly includes Spot, Scratch, Blister, Crack, Peeling, Cracking, Skinning, Icing and Greasy Dirt.

[0105] Optionally, the step S151 and the step S152 include:

[0106] The defect feature element of the blade defect picture is recognized by using a convolutional neural network pattern recognition model and a blade defect feature element comparison library.

[0107] The defect feature element of the blade defect picture and its type are marked out.

[0108] The defect feature element of the blade defect picture in the position corresponding to the blade defect picture is recognized by using a convolutional neural network digital recognition model.

[0109] In the specific embodiments of the present specification, the convolutional neural network pattern recognition model includes a CNN pattern recognition model. CNN is a kind of feedforward neural network, and the artificial neurons thereof can respond to a part of the surrounding units in the coverage range, which is very similar to the ordinary neural network, and they are both composed of neurons with learnable weights and bias constants. The default input of CNN is an image, and specific properties can be encoded into the network structure, so that the feedforward function is more efficient, and a large number of parameters are reduced. CNN is a kind of multi-layer neural network, which is good at processing image-related machine learning, especially large images. Convolutional network successfully reduces the dimension of the large amount of image recognition problem through a series of methods, and finally makes it possible to be trained. The hidden layer of CNN includes three common structures of convolution layer, pooling layer and full connection layer, wherein the convolution layer and the pooling layer cooperate to form multiple convolution groups, and the features are extracted layer by layer, and finally the classification is completed through a plurality of full connection layers. The operation completed by the convolution layer can be considered to be inspired by the concept of local receptive field, and the pooling layer is mainly used to reduce the data dimension. CNN simulates feature distinction through convolution, and reduces the order of magnitude of network parameters through convolution weight sharing and pooling.

[0110] The defect feature element of the blade defect picture is recognized by using a convolutional neural network pattern recognition model and combining a blade defect feature element comparison library. The defect feature element is circled on the blade picture with defects, and the type of the defect feature element of the blade defect picture is recognized by combining the blade defect feature element comparison library. Then, a convolutional neural network text recognition model is used to recognize the scale mark sprayed on the wind power blade, and the scale mark corresponding to the circled defect feature element is determined. The problem that the position of the defect feature element on the entire wind power blade cannot be accurately obtained due to the long length and high similarity of the wind power blade in the prior art is solved, the defect recognition and defect positioning are realized, and the efficiency and accuracy of defect recognition are improved.

[0111] S153: classify the corresponding blade defect picture based on the type of the defect feature element of the blade defect picture, and obtain a blade defect feature element comparison library.

[0112] In the specific embodiment of the present specification, the type of the defect feature element of the blade defect picture is taken as the defect type of the blade defect picture, and the defect pictures with the same defect type are classified to obtain a blade defect feature element comparison library, which is used for intelligent comparison of wind power blade defects.

[0113] S160: feed back the defect type comparison result to the user end.

[0114] In the specific embodiment of the present specification, the defect type comparison result is sent to the user end to remind the user end of the defect type of the wind power blade corresponding to the defect comparison video. The user end can adopt a mobile phone APP mode, so that the user end is not limited by the region and can receive the defect type of the wind power blade as long as the mobile phone has a signal; or the defect type of the wind power blade is transmitted through WIFI to realize real-time reception of the defect type of the wind power blade by the user end. The present application effectively solves the problems of easy omission in manual detection, difficult accurate recognition, short recognition and repair time caused by layer construction, and difficult comparison of historical quality retention of multi-layer paint surface.

[0115] In the specific embodiment of the present specification, a user uploads a leaf defect video, the leaf defect video is framed one by one to obtain a plurality of minimum granularity frame pictures, a convolutional neural network pattern recognition model is used to identify the defect feature elements of the frame pictures in combination with a leaf defect feature element comparison library. The defect feature elements are circled on the leaf pictures with defects, based on the matching results of the defect feature elements of the minimum granularity frame pictures and the leaf defect feature element comparison library constructed, a defect type comparison result is obtained, the defect type comparison result includes the defect type and the defect position, the defect type comparison result is sent to the user end to remind the user end of the defect type of the wind power generation blade corresponding to the defect comparison video. The user end can adopt a mobile phone APP mode, so that the user end is not limited by the region, as long as the mobile phone has a signal, the user end can accept the defect type of the wind power generation blade; or the WIFI transmission is used to realize the real-time reception of the defect type of the wind power generation blade by the user end. The present application effectively solves the problems of easy omission of artificial detection, difficult identification accuracy, short identification and repair time caused by layered construction, and difficult comparison of multi-layer paint history quality retention.

[0116] Figure 3 A principle schematic diagram of a wind power generation blade defect intelligent comparison device provided by the present embodiment is provided, which can include:

[0117] The acquisition module 10 is used to acquire a leaf defect video;

[0118] The framing module 20 is used to frame the leaf defect video one by one to obtain a plurality of minimum granularity frame pictures;

[0119] The identification module 30 is used to identify the defect feature elements in the minimum granularity frame pictures;

[0120] The marking module 40 is used to mark the positions of the defect feature elements of the minimum granularity frame pictures;

[0121] The comparison module 50 is used to match the defect feature elements of the minimum granularity frame pictures with a constructed leaf defect feature element comparison library to obtain a defect type comparison result;

[0122] The feedback module 60 is used to feed back the defect type comparison result to the user end.

[0123] Optionally, the constructed leaf defect feature element comparison library includes:

[0124] The defect feature elements of the existing leaf defect pictures are identified one by one;

[0125] The positions and types of the defect feature elements of the existing leaf defect pictures are marked;

[0126] According to the type of the defect feature element of the blade defect picture, the blade defect picture is classified, and a blade defect feature element comparison library is obtained.

[0127] Optionally, before the defect feature element of the blade defect picture is identified, the method further comprises:

[0128] A plurality of blade defect pictures to be processed are obtained.

[0129] It is judged whether the plurality of blade defect pictures to be processed have repeated pictures and / or invalid pictures.

[0130] When the plurality of blade defect pictures to be processed have repeated pictures and / or invalid pictures, the repeated pictures and / or invalid pictures are marked.

[0131] The marked blade defect pictures to be processed are removed, and existing blade defect pictures are obtained.

[0132] Optionally, the defect feature element of the obtained blade defect picture is identified, and the position and type of the defect feature element of the blade defect picture are marked, comprising:

[0133] The defect feature element of the blade defect picture is identified by using a convolutional neural network pattern recognition model and a blade defect feature element comparison library.

[0134] The defect feature element of the blade defect picture and its type are marked.

[0135] The defect feature element of the blade defect picture is identified in the corresponding position of the blade defect picture by using a convolutional neural network digital recognition model.

[0136] Optionally, before it is judged whether the plurality of blade defect pictures to be processed have repeated pictures and / or invalid pictures, the method further comprises: performing image processing on the plurality of blade defect pictures to be processed in at least one of the following modes: edge cutting, rotation, stretching, and image enhancement.

[0137] Optionally, the comparison module comprises:

[0138] A matching unit is configured to match the defect feature element of the minimum granularity frame picture with the constructed blade defect feature element comparison library, so as to obtain the similarity of the defect feature element of the minimum granularity frame picture and different defect types.

[0139] A judging unit is configured to judge whether the similarity of different defect types and the defect feature element of the minimum granularity frame picture exceeds a threshold value.

[0140] The sorting unit sorts all the defect types whose similarity to the defect feature elements of the minimum granularity sub-frame picture exceeds the threshold value, to obtain a defect type comparison result.

[0141] The functions of the device of the embodiments of the present application have been described in the above-mentioned method embodiments, and thus the description of the present embodiments will not be elaborated on the details.

[0142] Based on the same inventive concept, the embodiments of the present specification also provide an electronic device.

[0143] The electronic device embodiments of the present application are described below, which can be regarded as a specific physical implementation of the above-mentioned method and device embodiments of the present application. For the details described in the electronic device embodiments of the present application, it should be regarded as a supplement to the above-mentioned method or device embodiments; for the details not disclosed in the electronic device embodiments of the present application, it can be realized by referring to the above-mentioned method or device embodiments.

[0144] Figure 4 A structural schematic diagram of an electronic device provided by the embodiments of the present specification is shown. The electronic device 300 according to the embodiments of the present application is described below with reference to Figure 4 Figure 4 The electronic device 300 shown is merely an example and should not impose any limitation on the functions and use range of the embodiments of the present application.

[0145] As shown in Figure 4 The electronic device 300 is in the form of a general computing device. The components of the electronic device 300 can include but are not limited to: at least one processing unit 310, at least one storage unit 320, a bus 330 connecting different system components including the storage unit 320 and the processing unit 310, a display unit 340, etc.

[0146] The storage unit stores program codes which can be executed by the processing unit 310, so that the processing unit 310 performs the steps according to various exemplary embodiments of the present application described in the above processing method part of the present specification. For example, the processing unit 310 can perform the steps as shown in Figure 1

[0147] The storage unit 320 can include a readable medium in the form of a volatile storage unit, such as a random access memory (RAM) 3201 and / or a cache memory 3202, and can further include a read-only memory (ROM) 3203.

[0148] ​​The storage unit 320 can also include a program / utility 3204 having a set (at least one) of program modules 3205, including an operating system, one or more application programs, other program modules, and program data, each of which

[0149] The bus 330 can represent one or more of several types of bus structures, including a storage bus or bus controller, a peripheral bus, an accelerated graphics port, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MCA) bus, a Video Electronics Standards Association (VESA) local bus, and a proprietary bus implementing a variety of protocols that are frequently employed.

[0150] The electronic device 300 can also communicate with one or more external devices 400 such as a keyboard or pointing device, a Bluetooth device, etc.; other devices that enable a user to interact with the electronic device 300; and / or one or more devices that enable the electronic device 300 to communicate with one or more other computing devices. Such communication can be facilitated by an Input / Output (I / O) interface 350. Still yet, the electronic device 300 can communicate with one or more networks, such as a local area network (LAN), a general wide area network (WAN), and / or a public network such as the Internet, via a network adapter 360. The network adapter 360 can be communicatively coupled to the other components of the electronic device 300 via the bus 330. It should be appreciated that the bus 330 can be one or more busses, and can be implemented using any suitable type of architecture, including a bus architecture, a point-to-point architecture, a message switching architecture, etc. Figure 4 Other hardware and / or software modules that can be used in conjunction with the electronic device 300, such as microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data archival storage systems, etc. are not shown in FIG. 3 but can be incorporated into the electronic device 300.

[0151] Those skilled in the art will readily recognize that the example embodiments described with regard to the present application can be implemented using software and / or software in combination with necessary hardware. Thus, the technical solutions according to the embodiments of the present application can be embodied in the form of a software product, which can be stored in a computer readable storage medium (which can be a CD-ROM, a U disk, a mobile hard disk, etc.) or a network, and includes a plurality of instructions to enable a computing device (which can be a personal computer, a server, or a network device, etc.) to perform the above-described methods according to the present application. When the computer program is executed by a data processing device, the computer readable medium enables the computer readable medium to implement the above-described methods of the present application, i.e., the method shown in FIG. 4. Figure 1

[0152] Figure 5 A schematic diagram of a computer readable medium according to an embodiment of the present application.

[0153] Implementation​Figure 1 A computer program product of the method can be stored on one or more computer readable media. The computer readable media can be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium include an electrical connection having one or more wires, a portable disc, 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 disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0154] The computer readable storage medium can include a data signal transported over a carrier wave and can be contained in a baseband or propagated along with a carrier wave. The computer readable storage medium can also be distributed among many computers or devices. The computer readable storage medium can be in the form of packaged media or distributed media. The computer readable storage medium can be any available media or a combination of media that can be accessed by a general purpose or special purpose computer system. The computer readable storage medium can be a computer readable signal medium or a computer readable storage medium.

[0155] The program code can be executed by one or more programmable processors, which can be individually, or collectively, programmed to perform the operations described above. The program code can be implemented in any of various ways. For example, it can be implemented in any of various programming languages, including an object-oriented programming language such as Java, C++, or the like; and a conventional procedural programming language such as the "C" programming language or the like. The program code can execute entirely on a user's computing device, partly on the user's computing device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device or entirely on the remote computing device or server. In the latter scenario, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computing device, such as through the Internet using an Internet Service Provider (ISP). The program code can also be stored in a storage device or memory of a computing device.

[0156] In light of the above, the present application can be implemented in hardware, or implemented in software modules running on one or more processors, or implemented in a combination of the two. Those skilled in the art should understand that some or all of the functions of some or all of the components according to the embodiments of the present application can be implemented in practice using a general-purpose data processing device such as a microprocessor or a digital signal processor (DSP). The present application can also be implemented as a device or apparatus program (for example, a computer program and a computer program product) for performing part or all of the methods described herein. Such a program implementing the present application can be stored on a computer readable medium or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0157] The above-described specific embodiments further illustrate the purposes, technical solutions and beneficial effects of the present application, and it should be understood that the present application is not inherently related to any specific computer, virtual device or electronic device, and various general-purpose devices can also implement the present application. The above-described is only a specific embodiment of the present application, and is not used to limit the present application, and any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0158] Each of the embodiments in the specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment mainly describes the difference from other embodiments.

[0159] The above-described is only an embodiment of the present application, and is not used to limit the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the claims of the present application.

Claims

1. A wind power blade defect intelligent comparison method, characterized in that, The method comprises the following steps: acquiring a blade defect video; frame-by-frame dividing the blade defect video to obtain a plurality of minimum granularity frame pictures; identifying defect feature elements in the minimum granularity frame pictures; determining the positions of the defect feature elements on the blade surface according to scale marks sprayed on the surface of the wind power blade and having color difference with the blade base color; matching the defect feature elements of the minimum granularity frame pictures with a constructed blade defect feature element matching library to obtain a defect type matching result; wherein, the construction of the blade defect feature element matching library comprises: identifying defect feature elements of existing blade defect pictures one by one; marking the positions and types of the defect feature elements of the existing blade defect pictures according to the scale marks; classifying the existing blade defect pictures according to the types of the defect feature elements of the blade defect pictures to obtain the blade defect feature element matching library; feeding back the defect type matching result to a user terminal.

2. The wind power blade defect intelligent comparison method of claim 1, wherein, Before the step of identifying the defect feature elements of the blade defect pictures one by one, the method further comprises the following steps: acquiring a plurality of blade defect pictures to be processed; judging whether the plurality of blade defect pictures to be processed contain repeated pictures and / or invalid pictures; when the plurality of blade defect pictures to be processed contain repeated pictures and / or invalid pictures, marking the repeated pictures and / or invalid pictures; removing the marked blade defect pictures to be processed to obtain existing blade defect pictures. 3.The wind power blade defect intelligent comparison method of claim 2, wherein, The steps of identifying the defect feature elements of the acquired blade defect pictures one by one and marking the positions and types of the defect feature elements of the blade defect pictures comprise the following steps: identifying the defect feature elements of the blade defect pictures by using a convolutional neural network pattern recognition model and a blade defect feature element matching library; marking the defect feature elements of the blade defect pictures and their types; identifying the positions of the defect feature elements of the blade defect pictures in the corresponding blade defect pictures by using a convolutional neural network digital recognition model.

4. The wind power blade defect intelligent comparison method of claim 3, wherein, Before the step of judging whether the plurality of blade defect pictures to be processed contain repeated pictures and / or invalid pictures, the method further comprises the following step: performing image processing on the plurality of blade defect pictures to be processed by at least one of edge cutting, rotation, stretching and image enhancement. 5.The wind power blade defect intelligent comparison method of claim 1, wherein, The step of matching the defect feature elements of the minimum granularity frame pictures with the constructed blade defect feature element matching library to obtain a defect type matching result comprises the following steps: matching the defect feature elements of the minimum granularity frame pictures with the constructed blade defect feature element matching library to obtain the similarity of the defect feature elements of the minimum granularity frame pictures to different defect types; judging whether the similarity of different defect types to the defect feature elements of the minimum granularity frame pictures exceeds a threshold value; when the similarity of different defect types to the defect feature elements of the minimum granularity frame pictures exceeds the threshold value, sorting all the defect types with the similarity exceeding the threshold value to obtain a defect type matching result.

6. A wind power blade defect intelligent comparison device, characterized in that, The method comprises the following steps: an acquisition module is configured to acquire a blade defect video; a frame dividing module is configured to frame divide the blade defect video to obtain a plurality of minimum granularity frame pictures; The identification module is configured to identify a defect feature element in the minimum granularity subframe picture; The marking module is configured to determine the position of the defect feature element on the surface of the blade according to the scale mark sprayed on the surface of the wind power blade and having a color difference with the blade base color; The comparison module is configured to compare the defect feature element in the minimum granularity subframe picture with the constructed blade defect feature element comparison library to obtain a defect type comparison result; wherein the construction of the blade defect feature element comparison library comprises: identifying defect feature elements of existing blade defect pictures one by one; marking the position and type of the defect feature elements of the existing blade defect pictures according to the scale mark; classifying the corresponding blade defect pictures based on the type of the defect feature elements of the blade defect pictures to obtain a blade defect feature element comparison library; The feedback module is configured to feed back the defect type comparison result to a user end.

7. The wind power blade defect intelligent comparison device according to claim 6, wherein, Before the defect feature elements of the blade defect pictures are identified one by one, the method further comprises: obtaining a plurality of blade defect pictures to be processed; judging whether the plurality of blade defect pictures to be processed contain duplicate pictures and / or invalid pictures; when the plurality of blade defect pictures to be processed contain duplicate pictures and / or invalid pictures, marking the duplicate pictures and / or invalid pictures; removing the marked blade defect pictures to be processed to obtain existing blade defect pictures. 8.The wind power blade defect intelligent comparison device of claim 7, wherein, The identification of the defect feature elements of the obtained blade defect pictures, the marking of the position and type of the defect feature elements of the blade defect pictures, comprises: identifying the defect feature elements of the blade defect pictures by using a convolutional neural network pattern recognition model and a blade defect feature element comparison library; marking the defect feature elements of the blade defect pictures and their types; identifying the defect feature elements of the blade defect pictures in the corresponding positions of the blade defect pictures by using a convolutional neural network digital recognition model. 9.The wind power blade defect intelligent comparison device of claim 8, wherein, Before the judgment of whether the plurality of blade defect pictures to be processed contain duplicate pictures and / or invalid pictures, the method further comprises: performing image processing on the plurality of blade defect pictures to be processed in at least one of the following modes: edge cutting, rotation, stretching, and image enhancement. 10.The wind power blade defect intelligent comparison device of claim 9, wherein, The comparison module comprises: a matching unit configured to match the defect feature elements of the minimum granularity subframe picture with the constructed blade defect feature element comparison library to obtain the similarity of the defect feature elements of the minimum granularity subframe picture with different defect types; a judgment unit configured to judge whether the similarity of different defect types with the defect feature elements of the minimum granularity subframe picture exceeds a threshold value; a sorting unit configured to sort all the defect types whose similarity exceeds the threshold value when the similarity of different defect types with the defect feature elements of the minimum granularity subframe picture exceeds the threshold value to obtain a defect type comparison result.

11. An electronic device, comprising: The electronic device comprises: a processor; and a memory storing computer-executable instructions that, when executed, cause the processor to perform the method according to any one of claims 1-5. The electronic device comprises: a processor; and a memory storing computer-executable instructions that, when executed, cause the processor to perform the method according to any one of claims 1-5.

12. A computer readable storage medium, wherein, The computer-readable storage medium stores one or more programs, which when executed by the processor, implement the method of any one of claims 1-5.

Images