Distribution network unmanned aerial vehicle defect identification analysis method and system

Abstract

The invention discloses a distribution network unmanned aerial vehicle defect identification analysis method and system, and belongs to the technical field of defect identification and unmanned aerial vehicles, and the method comprises the steps: transforming a camera of an unmanned aerial vehicle, automatically focusing equipment, and carrying out the multi-angle shooting of the equipment after the unmanned aerial vehicle identifies a target region, and forming a line patrol picture library; processing the shot pictures to form a high-quality line patrol picture library, carrying out multi-angle shooting on normal lines and equipment to form a standard equipment picture library, and carrying out classification and identification according to regulations to form a defect data picture library; and comparing the line patrol picture library with a standard equipment picture library by using an image recognition algorithm in Matlab, marking different pictures, and carrying out one-to-many comparison on the different pictures and defect pictures in a defect data picture library so as to carry out defect recognition. According to the method, the fault range is quickly locked, the fault finding time is shortened, the quick power recovery efficiency is effectively improved, personal safety risks occurring in manual investigation are reduced to a great extent, the working efficiency is improved, and the working efficiency and the working quality are improved.

Description

Technical Field

[0001] The present invention relates to the field of defect identification and unmanned aerial vehicle technology, and in particular to a distribution network unmanned aerial vehicle defect identification and analysis method and system. Background Art

[0002] In recent years, with the continuous development of image processing technology and artificial intelligence algorithms, drones have been increasingly used in distribution network defect identification. However, the existing drone inspection technology still has certain limitations.

[0003] On the one hand, the existing drone cameras often have difficulty in achieving high-precision automatic focusing on lines and equipment during inspections, resulting in low-quality line inspection images, which affects the accuracy and efficiency of subsequent image processing. In addition, existing image preprocessing methods still have problems with poor recognition effects in terms of contrast enhancement, edge detection, and target area segmentation. These problems limit the application of drones in distribution network defect identification.

[0004] On the other hand, the existing technology lacks comprehensive consideration of multi-factor and multi-dimensional lines in terms of building standard equipment libraries and defect data libraries, such as different weather, light, and dust accumulation. In terms of image recognition algorithms, although there have been some research results, they have not yet been fully applied to the field of drone inspections, resulting in the need to improve the accuracy and efficiency of defect recognition. Summary of the invention

[0005] In view of the above-mentioned existing problems, the present invention proposes a method for distribution network drone defect identification and analysis to address the technical problems existing in the existing distribution network, such as poor drone inspection effect, low work efficiency, large amount of manpower required for defect identification, long time for defect entry into the system, and high professional and technical requirements for personnel. By transforming the drone camera, optimizing the image processing process and adopting advanced image recognition algorithms, it is expected to significantly improve the accuracy and efficiency of distribution network defect identification.

[0006] In order to solve the above technical problems, a distribution network UAV defect identification and analysis method is proposed, including:

[0007] The camera of the drone is modified to automatically focus the equipment. After the drone identifies the target area, it shoots the equipment from multiple angles to form a line inspection picture library. The captured pictures are processed to form a high-quality line inspection picture library. Normal lines and equipment are photographed from multiple angles to form a standard equipment library, which is classified and identified according to regulations to form a defect data library. The image recognition algorithm in Matlab is used to compare the line inspection picture library with the standard equipment library, annotate the difference pictures, and perform a one-to-many comparison with the defect pictures in the defect data library to identify defects.

[0008] As a preferred solution of the defect identification and analysis method of a distribution network drone described in the present invention, the line inspection picture library includes: transforming the camera of the drone into a device that automatically focuses on the line, hardware, insulators, and lightning arresters, so that the camera is focused on the line and equipment, effectively distinguishing the line or equipment in the camera from the background, performing image preprocessing, and performing contrast enhancement;

[0009] Use the enhanced image for edge detection. When the pixel value of the image after contrast enhancement is greater than μ local +k·σ local When the edge detection binary output is 1, when the image pixel value after contrast enhancement does not exceed μ local +k·σ local When , the binary output of edge detection is 0;

[0010] Segment the target area and the background part, calculate the importance of the current area in the segmentation process, and when the importance exceeds the first threshold, it is considered to be the target area of ​​the line or device. When the importance is less than the first threshold, it is considered to be the background part for segmentation and hiding, and quickly focus on the target area of ​​the line or device in the camera, and focus on the target area;

[0011] After the drone identifies the line or equipment area, it takes photos of the line and equipment from multiple angles to form a line inspection picture library for this line inspection.

[0012] As a preferred solution of the defect identification and analysis method of a distribution network drone described in the present invention, the high-quality line patrol picture library includes removing some influencing factors by edge detection, filtering operation, binarization, image scaling, defogging, and photo weighting of the pictures in the line patrol picture library in environments with dim lighting, excessive lighting, and low clarity, and then performing linear correction, noise removal, bad pixel removal, interpolation, white balance, and automatic exposure control to restore picture details to form a high-quality line patrol picture library;

[0013] The processed image has some noise in a certain period of time or a certain frame, which will cause some interference in image recognition. The binary image E'(x,y) is obtained by binarizing the image:

[0014]

[0015] Among them, E(x, y) is the original image, T is the noise reduction threshold, and the open-loop closed operation of morphological filtering is used to perform noise reduction on the binary processed metal micro-position motion image to remove noise points unrelated to the moving object:

[0016]

[0017] Where N is the working space of the moving object, B is the structural element, ° is the morphological opening operation, and · is the morphological closing operation. is the erosion operation, and ⊕ is the dilation operation.

[0018] As a preferred solution of the defect identification and analysis method of a distribution network drone described in the present invention, the standard equipment library includes: normal lines and equipment including those without heating, discharge, and flashover phenomena are photographed at multiple angles to form a standard equipment library;

[0019] The defect data library includes constructing a defect data library of multi-factor and multi-dimensional lines, wherein the multi-factors include different weather, different light, and different dust accumulation levels, and the multi-dimensional lines include pictures of defects such as line aging, line breakage, line burrs, lightning strike marks, insulator damage, and hardware corrosion across regions, years, and structures to form a defect data library.

[0020] As a preferred solution of the defect identification and analysis method of a distribution network drone described in the present invention, the image recognition algorithm includes performing image recognition and comparison according to the Thiessen polygon feature decomposition method, and in the target distribution network area, s={s1,s2,…s n} represents a group of sensor nodes, each node corresponds to a device or line component that may have defects. An effective monitoring network is established through the nodes. The entire area is divided into n convex polygons through the Thiessen polygon algorithm, so that each polygon contains only one node, which is expressed as:

[0021] Π i = {q∈R 2 |d(q,s i )≤d(q,s j ),j∈{1,…,n},j≠i}

[0022] Among them, i and j are variable indexes, q is an arbitrary point, and s i and j are any two adjacent nodes;

[0023] The device or line feature vectors of the pictures taken by the drone line patrol include the shape, color, and texture of the equipment, and the feature vectors of the pictures taken by the drone line patrol are compared with the line and equipment standard database for equipment and line identification.

[0024] As a preferred solution of the distribution network drone defect identification and analysis method described in the present invention, the comparison includes comparing the pictures in the line patrol picture library each time with the pictures in the standard equipment picture library through the algorithm in Matlab, extracting the equipment or line feature vectors of the pictures taken by the drone line patrol, calculating the relative difference measure with the standard feature vector, and when the relative difference measure is greater than 0.85, judging the differences in the input pictures, and marking the equipment status, location and defect type, and automatically filtering the identified difference pictures into the difference picture folder.

[0025] As a preferred solution of the defect identification and analysis method for a distribution network drone described in the present invention, the comparison further includes setting a comparison threshold of the screened pictures with differences and the defect pictures in the defect data library to 60%, and when the contrast ratio between the screened pictures with differences and the defect pictures in the defect data library is less than 60%, no comparison is performed;

[0026] When the contrast ratio between the screened out pictures with differences and the defect pictures in the defect data library exceeds 60%, the defect pictures of the same type of lines or equipment in the defect data library are retrieved for one-to-many comparison to determine the multiple defects and multiple anomalies of the same equipment defect.

[0027] Another object of the present invention is to provide a distribution network drone defect identification and analysis system, which aims to solve technical problems in drone inspections, including improving the accuracy of image acquisition, ensuring the acquisition of clear inspection images through camera modification and image preprocessing; improving the efficiency and accuracy of defect identification, using image recognition algorithms to automatically compare standard equipment libraries and defect data libraries to quickly locate defects; handling the diversity and complexity of defect types, and achieving accurate classification by building a comprehensive defect data library; and optimizing the output and application of defect identification results, providing convenient defect information management for power grid operation and maintenance by screening, labeling and classifying difference images, thereby comprehensively improving the efficiency and safety of power grid operation and maintenance.

[0028] As a preferred solution of the network distribution drone defect identification and analysis system described in the present invention, it is characterized by comprising an image acquisition module, a defect comparison module, and a defect identification module;

[0029] The image acquisition module is responsible for transforming the drone camera to achieve automatic focusing, taking multi-angle photos of the target area, generating a line patrol picture library, and pre-processing the taken pictures to form a high-quality line patrol picture library;

[0030] The defect comparison module constructs a standard equipment library and a defect data library, and uses an image recognition algorithm to compare the line inspection image library with the standard equipment library, annotate the difference images, and then perform a one-to-many comparison with the defect data library;

[0031] The defect recognition module screens the identified difference images, marks the equipment status, location and defect type, and outputs the results to a difference image folder.

[0032] A computer device includes a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of the method for identifying and analyzing defects of a distribution network drone are implemented.

[0033] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the method for identifying and analyzing defects of a distribution network drone are implemented.

[0034] Beneficial effects of the present invention: Advantages of the defect identification method of the unmanned aerial vehicle of the present invention in the actual application process: In terms of fault finding, combined with the action of the automatic switch, the fault range is quickly locked, the process of climbing mountains and crossing rivers for finding is eliminated, the fault finding time is reduced, and the efficiency of rapid power restoration is effectively improved. In terms of disaster area survey: for disasters such as floods and wildfires, unmanned aerial vehicles are used to enter the disaster area for multiple round trips to conduct surveys, collect the on-site environment of the disaster area, the operation and damage of line equipment, and provide accurate disaster information in time for emergency rescue and power restoration. Greatly reduce the personal safety risks in manual surveys and improve work efficiency. Use unmanned aerial vehicles to conduct on-site high-altitude aerial photography of traffic road expansion and municipal construction power relocation projects, substation preliminary site selection, 10 kV wiring reconstruction projects, etc., to more clearly understand the on-site environment of regional cross-roads, cross-rivers and adjacent cross-crossings, operating equipment, etc., thereby improving the efficiency of on-site surveys, effectively solving the problems of running by car, searching on-site, not being able to go, and not being able to see all, and providing strong picture data support for the design and line selection of power construction projects and on-site surveys and construction plan preparation of projects under construction, and improving work efficiency and work quality. During the completion and intermediate inspection of new line construction and relocation projects, drones are used to conduct aerial photography inspections of key areas such as pole tower assembly, hardware cross arm installation, and line binding. This can replace the previous manual pole climbing inspection and acceptance, reduce the risks of pole climbing operations, improve acceptance efficiency and quality, and use pictures as evidence to ensure "zero defects" when the line is put into operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work, among which:

[0036] Figure 1An overall flow chart of a method for identifying and analyzing network distribution drone defects provided by an embodiment of the present invention.

[0037] Figure 2 A Thiessen polygon algorithm diagram of a distribution network drone defect identification and analysis method provided by an embodiment of the present invention.

[0038] Figure 3 A schematic diagram of the module structure of a distribution network drone defect identification and analysis system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0039] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0040] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0041] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it an embodiment that is mutually exclusive with other embodiments, either individually or selectively.

[0042] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0043] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0044] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0045] Example 1, reference Figure 1-Figure 2 , which is the first embodiment of the present invention, and provides a method for identifying and analyzing defects of a distribution network UAV, comprising:

[0046] S1: The camera of the drone is modified to automatically focus the device. After the drone identifies the target area, it takes multi-angle photos of the device to form a line inspection picture library.

[0047] It should be noted that the camera of the drone is transformed into a device that automatically focuses on the line, hardware, insulators, and lightning arresters, so that the camera is focused on the line and equipment, and the line or equipment in the camera is effectively distinguished from the background part, and image preprocessing and contrast enhancement are performed:

[0048]

[0049] Among them, E(x,y) represents the pixel value of the image after contrast enhancement, I(x,y) is the value of the pixel in the original image, and μ global is the global mean of all pixel values ​​in the original image, σ local is the standard deviation of the original image in the detection area, a is a constant to prevent the denominator from being zero, a=0.01;

[0050] Use the enhanced image for edge detection. When the pixel value E(x,y) of the image after contrast enhancement is greater than μ local +k·σ local When the edge is detected, the binary output B(x,y) is 1, and when the image pixel value E(x,y) after contrast enhancement does not exceed μ local +k·σ local When , the edge detection binary output B(x,y) is 0, where k is the adjustment parameter;

[0051] Segment the target area and the background:

[0052]

[0053] Among them, A represents the size of the space occupied by the detected area, β represents the adjustment coefficient, and S represents the importance of the current area in the segmentation process;

[0054] When S exceeds the first threshold, it is considered to be the target area of ​​the line or device. When S is less than the first threshold, it is considered to be the background part and segmented and hidden. The camera is quickly focused on the target area of ​​the line or device. Focus is performed on the target area:

[0055]

[0056] Among them, D represents the focal length of the lens, C max is the maximum limit of visible depth, f max Indicates the maximum focus range available for the device;

[0057] After the drone identifies the line or equipment area, it takes photos of the line and equipment from multiple angles to form a line inspection picture library for this line inspection.

[0058] S2: Process the captured images to form a high-quality line inspection image library. Photograph normal lines and equipment from multiple angles to form a standard equipment image library. Classify and identify defects according to regulations to form a defect data image library.

[0059] It should be noted that the images of the line patrol image library are subjected to edge detection, filtering, binarization, image scaling, defogging, and photo weighting in environments with dim light, strong light, and low definition to remove some influencing factors, and then linear correction, noise removal, bad pixel removal, interpolation, white balance, and automatic exposure control are performed to restore image details, thus forming a high-quality line patrol image library;

[0060] The processed image has some noise in a certain period of time or a certain frame, which will cause some interference in image recognition. The binary image E'(x,y) is obtained by binarizing the image:

[0061]

[0062] Among them, E(x, y) is the original image, T is the noise reduction threshold, and the open-loop closed operation of morphological filtering is used to perform noise reduction on the binary processed metal micro-position motion image to remove noise points unrelated to the moving object:

[0063]

[0064] Where N is the working space of the moving object, B is the structural element, ° is the morphological opening operation, and · is the morphological closing operation. is the erosion operation, and ⊕ is the dilation operation.

[0065] The specific code is as follows:

[0066]

[0067] The line inspection images are subjected to edge detection, filtering, binarization, image scaling, defogging, and photo weighting to remove some influencing factors. Then linear correction, noise removal, bad pixel removal, interpolation, white balance, automatic exposure control and other measures are performed to restore image details, greatly improving the quality of the images.

[0068] It is understandable that the standard equipment library includes taking multi-angle photos of normal lines and equipment including those without heating, discharge, and flashover to form a standard equipment library;

[0069] The defect data library includes constructing a defect data library of multi-factor and multi-dimensional lines, wherein the multi-factors include different weather, different light, and different dust accumulation levels, and the multi-dimensional lines include pictures of defects such as line aging, line breakage, line burrs, lightning strike marks, insulator damage, and hardware corrosion across regions, years, and structures to form a defect data library.

[0070] S3: Use the image recognition algorithm in Matlab to compare the line inspection image library with the standard equipment image library, mark the difference images, and perform a one-to-many comparison with the defect images in the defect data library to identify defects.

[0071] It should be noted that if Figure 2 As shown in the figure, according to the Thiessen polygon feature decomposition method, image recognition and comparison are performed. In the target distribution network area, s={s1,s2,…s n} represents a group of sensor nodes, each node corresponds to a device or line component that may have defects. An effective monitoring network is established through the nodes. The entire area is divided into n convex polygons through the Thiessen polygon algorithm, so that each polygon contains only one node, which is expressed as:

[0072] Π i = {q∈R 2 |d(q,s i )≤d(q,s j ),j∈{1,…,n},j≠i}

[0073] Among them, i and j are variable indexes, q is an arbitrary point, and s i and j are any two adjacent nodes;

[0074] Extract the device or line feature vectors from the pictures taken by the drone, including the shape, color, and texture of the equipment, and compare the feature vectors of the pictures taken by the drone with the line and equipment standard database for equipment and line identification;

[0075] Specifically, the pictures in each line inspection picture library are compared with the pictures in the standard equipment picture library through the algorithm in Matlab, and the equipment or line feature vectors of the pictures taken by the drone line inspection are extracted by nonlinear transformation to obtain F transformed , calculate the standard eigenvector F standard Relative difference measure of :

[0076]

[0077] When Sim relative If the error is greater than 0.85, the input images are judged to be different, and the equipment status, location and defect type are marked. The identified difference images are automatically filtered into the difference image folder.

[0078] Furthermore, the threshold for comparing the images screened out with differences with the defect images in the defect data library is set to 60%, and when the contrast ratio between the images screened out with differences and the defect images in the defect data library is less than 60%, no comparison is performed;

[0079] When the contrast ratio between the screened out images with differences and the defect images in the defect data library exceeds 60%, the defect images of the same type of lines or equipment in the defect data library are retrieved for one-to-many comparison to determine the multiple defects and multiple anomalies of the same equipment defect; thus greatly improving the time and accuracy of defect identification.

[0080] Embodiment 2, the second embodiment of the present invention, is different from the previous embodiment in that:

[0081] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program codes.

[0082] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0083] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0084] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0085] Example 3, reference Figure 3 , which is the third embodiment of the present invention, and provides a network distribution drone defect recognition and analysis system, including an image acquisition module 10, a defect comparison module 20, and a defect recognition module 30;

[0086] The image acquisition module 10 is responsible for transforming the drone camera to achieve automatic focusing, and taking multi-angle photos of the target area to generate a line patrol picture library, and pre-processing the taken pictures to form a high-quality line patrol picture library;

[0087] The defect comparison module 20 constructs a standard equipment library and a defect data library, and uses an image recognition algorithm to compare the line inspection picture library with the standard equipment library, annotate the difference pictures, and then perform a one-to-many comparison with the defect data library;

[0088] The defect recognition module 30 screens the identified difference pictures, marks the equipment status, location and defect type, and outputs the results to a difference picture folder.

[0089] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for identifying and analyzing defects of a distribution network drone, characterized in that: include, The drone's camera is modified to automatically focus the device. After the drone identifies the target area, it takes photos of the device from multiple angles to form a line inspection picture library. Process the captured images to form a high-quality line inspection image library, take multi-angle photos of normal lines and equipment to form a standard equipment image library, and classify and identify defects according to regulations to form a defect data image library; Use the image recognition algorithm in Matlab to compare the line inspection image library with the standard equipment image library, annotate the difference images, and perform a one-to-many comparison with the defect images in the defect data library to identify defects.

2. A method for identifying and analyzing network distribution drone defects as described in claim 1, characterized in that: The line inspection image library includes transforming the camera of the drone into a device that automatically focuses on the line, hardware, insulators, and lightning arresters, so that the camera is focused on the line and equipment, effectively distinguishing the line or equipment in the camera from the background, performing image preprocessing, and performing contrast enhancement; Use the enhanced image for edge detection. When the pixel value of the image after contrast enhancement is greater than μ local +k·σ local When the edge detection binary output is 1, when the image pixel value after contrast enhancement does not exceed μ local +k·σ local When , the binary output of edge detection is 0; Segment the target area and the background part, calculate the importance of the current area in the segmentation process, and when the importance exceeds the first threshold, it is considered to be the target area of ​​the line or device. When the importance is less than the first threshold, it is considered to be the background part for segmentation and hiding, and quickly focus on the target area of ​​the line or device in the camera, and focus on the target area; After the drone identifies the line or equipment area, it takes photos of the line and equipment from multiple angles to form a line inspection picture library for this line inspection.

3. A method for identifying and analyzing network distribution drone defects as described in claim 2, characterized in that: The high-quality line patrol picture library includes removing some influencing factors by edge detection, filtering, binarization, image scaling, defogging, and photo weighting in environments with dim light, strong light, and low definition, and then performing linear correction, noise removal, bad pixel removal, interpolation, white balance, and automatic exposure control to restore picture details to form a high-quality line patrol picture library; The processed image has some noise in a certain period of time or a certain frame, which will cause some interference in image recognition. The binary image E'(x,y) is obtained by binarizing the image: Among them, E(x, y) is the original image, T is the noise reduction threshold, and the open-loop closed operation of morphological filtering is used to perform noise reduction on the binary processed metal micro-position motion image to remove noise points unrelated to the moving object: Among them, N is the working space of the moving object, B is the structural element, is the morphological opening operation, is the morphological closing operation, is the corrosion operation, The expansion operation.

4. A method for identifying and analyzing network distribution drone defects as described in claim 3, characterized in that: The standard equipment library includes taking multi-angle photos of normal lines and equipment including those without heating, discharge, and flashover to form a standard equipment library; The defect data library includes constructing a defect data library of multi-factor and multi-dimensional lines, wherein the multi-factors include different weather, different light, and different dust accumulation levels, and the multi-dimensional lines include pictures of defects such as line aging, line breakage, line burrs, lightning strike marks, insulator damage, and hardware corrosion across regions, years, and structures to form a defect data library.

5. A method for identifying and analyzing network distribution drone defects as described in claim 4, characterized in that: The image recognition algorithm includes performing image recognition and comparison according to the Thiessen polygon feature decomposition method. In the target distribution network area, s={s1,s2,…s n } represents a group of sensor nodes, each node corresponds to a device or line component that may have defects. An effective monitoring network is established through the nodes. The entire area is divided into n convex polygons through the Thiessen polygon algorithm, so that each polygon contains only one node, which is expressed as: Π i ={q∈R 2 |d(q,s i )≤d(q,s j ),j∈{1,…,n},j≠i} Among them, i and j are variable indexes, q is an arbitrary point, and s i and j are any two adjacent nodes; The device or line feature vectors of the pictures taken by the drone line patrol include the shape, color, and texture of the equipment, and the feature vectors of the pictures taken by the drone line patrol are compared with the line and equipment standard database for equipment and line identification.

6. A method for identifying and analyzing network distribution drone defects as claimed in claim 5, characterized in that: The comparison includes comparing the pictures in each line patrol picture library with the pictures in the standard equipment picture library through the algorithm in Matlab, performing nonlinear transformation on the equipment or line feature vectors extracted from the pictures taken by the drone line patrol, calculating the relative difference measure with the standard feature vector, and when the relative difference measure is greater than 0.85, judging the differences between the input pictures, and marking the equipment status, location and defect type, and automatically filtering the identified difference pictures into the difference picture folder.

7. A method for identifying and analyzing network distribution drone defects as claimed in claim 6, characterized in that: The comparison further includes setting a comparison threshold of the screened pictures with differences and the defect pictures in the defect data library to 60%, and when the contrast ratio between the screened pictures with differences and the defect pictures in the defect data library is less than 60%, no comparison is performed; When the contrast ratio between the screened out pictures with differences and the defect pictures in the defect data library exceeds 60%, the defect pictures of the same type of lines or equipment in the defect data library are retrieved for one-to-many comparison to determine the multiple defects and multiple anomalies of the same equipment defect.

8. A system using a distribution network drone defect identification and analysis method as claimed in any one of claims 1 to 7, characterized in that: Including image acquisition module, defect comparison module and defect recognition module; The image acquisition module is responsible for transforming the drone camera to achieve automatic focusing, taking multi-angle photos of the target area, generating a line patrol picture library, and pre-processing the taken pictures to form a high-quality line patrol picture library; The defect comparison module constructs a standard equipment library and a defect data library, and uses an image recognition algorithm to compare the line inspection image library with the standard equipment library, annotate the difference images, and then perform a one-to-many comparison with the defect data library; The defect recognition module screens the identified difference images, marks the equipment status, location and defect type, and outputs the results to a difference image folder.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of a distribution network drone defect identification and analysis method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a distribution network drone defect identification and analysis method described in any one of claims 1 to 7 are implemented.

Images