Insulator defect sample generation method, system, device and computer-readable storage medium based on small amount of data

Abstract

The present invention provides a method, system, device and computer-readable storage medium for generating insulator defect samples based on a small amount of data, wherein the method for generating insulator defect samples based on a small amount of data includes dividing existing insulator data into first data containing defective insulators and second data not containing defective insulators; obtaining first defective insulator image data and first background image data in the first data through a target algorithm; obtaining first normal insulator image data and second background image data in the second data through a target algorithm; expanding insulator defect sample data; and combining the expanded insulator defect sample data with the first data to generate new insulator defect samples. Different defect generation strategies are adopted for insulator images with and without defects, and a large number of samples close to real data can be expanded in the case of very few samples.

Description

Technical Field

[0001] The present invention relates to the technical field of insulator defect detection in distribution networks, and in particular to a method, system, device and computer-readable storage medium for generating insulator defect samples based on a small amount of data. Background Art

[0002] With the development of artificial intelligence and the Internet of Things, existing technologies have enabled drones to detect insulators on distribution networks in real time, and accurately identify defects such as insulator self-explosion, damage, and contamination. Compared with traditional manual inspections of insulator defects, drones have the characteristics of fast detection speed, high accuracy, and high efficiency.

[0003] In actual production environments, it is often difficult to collect insulator defect data, which makes it difficult to apply many data-dependent deep learning methods. As a result, many insulator defect data generation technologies have emerged. Among them, the Generative Adversarial Network (Gan) is used in the generation of defect samples for power inspections. However, the current sample generation technology based on Gan relies on a large number of defect samples as a priori data sets and the generated amount is insufficient.

[0004] To solve the above problems, the present invention provides a method, system, device and computer-readable storage medium for generating insulator defect samples based on a small amount of data. Summary of the invention

[0005] In order to solve the problems in the prior art, the present invention provides a method, system, device and computer-readable storage medium for generating insulator defect samples based on a small amount of data.

[0006] The technical solution of the present invention is as follows:

[0007] In a first aspect of the present application, a method for generating insulator defect samples based on a small amount of data is provided, which specifically comprises the following steps:

[0008] The existing insulator data is divided into first data containing defective insulators and second data not containing defective insulators;

[0009] Acquire the first defective insulator image data and the first background image data in the first data by using a target algorithm;

[0010] Acquire the first normal insulator image data and the second background image data in the second data by using a target algorithm;

[0011] Expand the sample data of insulator defects:

[0012] Performing geometric transformation on each defective insulator image in the first defective insulator image data to form second defective insulator image data, and fusing each second insulator defect image in the second defective insulator image data with each background image in the first background image data to form a first expanded sample;

[0013] Perform defect simulation on each first normal insulator image in the first normal insulator image data to form first simulated defective insulator image data, and fuse each first simulated defective insulator image in the first simulated defective insulator image data with each second background image in the second background image data to form a second expanded sample;

[0014] The first data and the expanded insulator defect sample data are combined into a new insulator defect sample.

[0015] In a possible implementation of the first aspect, the expanded insulator defect sample data further includes a third expanded sample, and the third expanded sample further includes the following steps:

[0016] Acquire multiple natural scene pictures, and combine the natural scene picture sets into natural scene data;

[0017] Acquire second defective insulator image data, and fuse each second insulator defect image in the second defective insulator image data with each natural scene image in the natural scene data to form first simulation data;

[0018] Acquire first simulated defective insulator image data, and fuse each first simulated defective insulator image in the first simulated defective insulator image data with each natural scene image in the natural scene data to form second simulated data;

[0019] The first simulated data is combined with the second simulated data to form a third expanded sample.

[0020] In a possible implementation of the first aspect, detecting the accuracy of the expanded insulator defect sample data, detecting the accuracy of the expanded data sample specifically includes the following steps:

[0021] Using the first expanded sample as a defect detection set and the second expanded sample as a test data set;

[0022] Use the defect detection set to detect the accuracy of the test data set;

[0023] When the judgment accuracy is not less than the set threshold;

[0024] The expanded insulator defect sample data is incorporated into the new insulator defect sample.

[0025] In a possible implementation of the first aspect above, when the judgment accuracy is less than a set threshold;

[0026] The first expanded sample and the second expanded sample are regenerated.

[0027] In a possible implementation of the first aspect above, the geometric transformation includes flipping, rotation and translation.

[0028] In a possible implementation of the first aspect above, the defect simulation includes string drop, breakage, shed contamination and self-explosion.

[0029] The second aspect of the present application provides a method for generating insulator defect samples based on a small amount of data, which is applied to the method for generating insulator defect samples based on a small amount of data provided in the first aspect, comprising:

[0030] A target detection module, used for dividing the existing insulator data into first data containing defective insulators and second data not containing defective insulators;

[0031] A separation module, used for separating the first defective insulator image data and the first background image data in the first data;

[0032] and separating the first normal insulator image data and the second background image data in the second data;

[0033] The expansion module is used to convert the first data into a first expanded sample and the second data into a second expanded sample, and to form a new insulator defect sample by combining the first data, the first expanded sample and the second expanded sample.

[0034] In a possible implementation of the second aspect, a supplementing module is further included, wherein the supplementing module is used to generate a third expanded sample, and the third expanded sample is used to expand the data volume of the new insulator defect sample.

[0035] In a possible implementation of the second aspect, an evaluation module is further included. The evaluation module uses the first expanded sample as a defect detection set, the second expanded sample as a test data set, and uses the defect detection set to detect the accuracy of the test data set.

[0036] A third aspect of the present application provides an insulator defect sample device based on a small amount of data, comprising:

[0037] Memory for storing computer programs;

[0038] A processor is used to implement a method for generating insulator defect samples based on a small amount of data when executing the computer program.

[0039] A fourth aspect of the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, a method for generating an insulator defect sample based on a small amount of data is implemented.

[0040] Compared with the prior art, the technical solution proposed by the present invention has the following beneficial effects:

[0041] Different defect generation strategies are adopted for insulator images with defects and without defects. With very few samples, a large number of samples close to real data can be expanded.

[0042] Different image processing strategies are used for different defect types, and the defect generation details can be adjusted to be close to the real data and improve the realism.

[0043] The present invention uses the first expanded sample as a defect detection set and the second expanded sample as a test data set, and uses the defect detection set to detect the accuracy of the test data set, so as to iteratively optimize the quality of the generated defect data set. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0045] Figure 1 According to an embodiment of the present application, a flow chart of a method for generating insulator defect samples based on a small amount of data is shown;

[0046] Figure 2 According to an embodiment of the present application, a flow chart of a method for generating expanded insulator defect sample data is shown;

[0047] Figure 3 According to an embodiment of the present application, a schematic flow chart of a third method for generating an expanded sample is shown;

[0048] Figure 4 According to an embodiment of the present application, a schematic diagram of dropped string generation based on a skeleton method is shown;

[0049] Figure 5 According to an embodiment of the present application, a schematic diagram of a target algorithm is shown;

[0050] Figure 6a According to an embodiment of the present application, a schematic diagram of a pre-selected ROI region in a first normal insulator image is shown;

[0051] Figure 6b According to an embodiment of the present application, a schematic diagram of generating a dirty material image is shown;

[0052] Figure 6cAccording to an embodiment of the present application, a schematic diagram of fusion-generated umbrella skirt pollution samples is shown;

[0053] Figure 7 According to an embodiment of the present application, a flowchart of actual image changes of a drop string generation based on a skeleton method is shown;

[0054] Figure 8 According to an embodiment of the present application, a structural schematic diagram of an insulator defect sample production system based on a small amount of data is shown;

[0055] Fig. 9 According to an embodiment of the present application, a schematic structural diagram of an insulator defect sample device based on a small amount of data is shown;

[0056] Fig.10 According to an embodiment of the present application, a schematic diagram of the structure of a computer-readable storage medium is shown. DETAILED DESCRIPTION

[0057] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0058] As used herein, the term "including" and its variations mean open inclusion, i.e., "including but not limited to". Unless otherwise stated, the term "or" means "and / or". The term "based on" means "based at least in part". The terms "an example embodiment" and "an embodiment" mean "at least one example embodiment". The term "another embodiment" means "at least one additional embodiment". The terms "first", "second", etc. may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0059] In view of the technical problems existing in the prior art, the present application provides a method, system, device and computer-readable storage medium for generating insulator defect samples based on a small amount of data. The technical solution provided by the present application will be explained and illustrated in combination with embodiments below.

[0060] In some embodiments of the present application, Figure 1-7 As shown, a method for generating insulator defect samples based on a small amount of data specifically includes the following steps:

[0061] S100: dividing the existing insulator data into first data containing defective insulators and second data not containing defective insulators;

[0062] S200: Acquire first defective insulator image data and first background image data in the first data by using a target algorithm;

[0063] S300: Acquire first normal insulator image data and second background image data in the second data by using a target algorithm;

[0064] Figure 5 According to an embodiment of the present application, a schematic diagram of a target algorithm is shown, insulator data containing defects is processed, defects including string drop, breakage, shed contamination, etc. of the insulator are detected, and the data is divided into defective and non-defective samples through the classification of defects by the target detection algorithm. The target algorithm in this embodiment adopts the GrabCut algorithm, which is used to separate each defective insulator image in the first data into a first defective insulator image and a first background image, all first defective insulator images constitute the first defective insulator image data, and the first background image constitutes the first background image data; each non-defective insulator image in the second data is separated into a first normal insulator image and a second background image, all first normal insulator images constitute the first normal insulator image data, and the second background image constitutes the second background image data.

[0065] S400: Expanded insulator defect sample data: Figure 2 According to an embodiment of the present application, a flow chart of a method for generating expanded insulator defect sample data is shown;

[0066] S410: geometrically transform each defective insulator image in the first defective insulator image data to form second defective insulator image data, and fuse each second insulator defect image in the second defective insulator image data with each background image in the first background image data to form a first expanded sample;

[0067] S420: Perform defect simulation on each first normal insulator image in the first normal insulator image data to form first simulated defective insulator image data, and fuse each first simulated defective insulator image in the first simulated defective insulator image data with each second background image in the second background image data to form a second expanded sample.

[0068] In this embodiment, for an insulator image without defects, in the insulator image with the background separated, the insulator ROI area where defects are to be generated is manually selected. To ensure the authenticity of defect generation, the area is analyzed and the defect type is independently selected. The defect types include dropped strings, breakage, shed contamination, and self-explosion; the geometric transformation includes flipping, rotation, and translation; defects are simulated and generated, and the defects are merged into the second background image in the randomly selected second background image data.

[0069] In a specific example of this embodiment, a ROI area is pre-selected in the first normal insulator image, defect simulation is performed according to defect types including string drop, breakage, shed contamination, and self-explosion, and a first simulated defective insulator image is generated, and finally the first simulated defective insulator image is fused with two arbitrarily selected background images to form data in the second expanded sample.

[0070] Figure 6a-6c According to an embodiment of the present application, the process of generating a shed pollution sample in the second extended sample is shown; see Figure 6a As shown, the ROI area is preferably pre-selected in the first normal insulator image, see Figure 6b As shown, for the umbrella skirt pollution, the pollution material is randomly selected to generate the pollution material image, see Figure 6c As shown, the first simulated defective insulator image is fused with two randomly selected background images to generate a shed contamination sample. Contamination brightness adjustment and edge processing, first convert the contamination map and ROI area map into HSV format, calculate the average brightness of the non-black area, and align the brightness of the contamination map to the average brightness level of the ROI area.

[0071] Figure 4 According to an embodiment of the present application, a schematic diagram of a dropped string generation based on a skeleton method is shown. In this embodiment, the dropped string generation based on the skeleton method includes the following specific steps:

[0072] Step S10: reading an image to obtain a first normal insulator image in the second data;

[0073] Step S20: image binarization, binarizing the first normal insulator image in step S10. Image binarization simplifies the first normal insulator image, reduces the amount of data, and highlights the outline of the normal insulator.

[0074] Step S30: thinning the binary image and extracting the skeleton. Based on step S30, the edge of the binary image of the first normal insulator image is thinned, and then the skeleton image of the normal insulator is extracted.

[0075] Step S40: performing restricted region growing to improve the clarity of the skeleton image.

[0076] Step S50: Delete the corresponding skeleton points according to the selected roi area; delete the skeleton points corresponding to the selected string drop area in the normal insulator.

[0077] Step S60: removing useless discrete points of the original image according to the skeleton points.

[0078] Step S70: post-processing such as edge processing; improving the first simulated defective insulator image after the string drop processing.

[0079] Step S80: merging the defective insulator image into the background; the first simulated defective insulator image is merged with each second background image in the second background image data to form data in the second expanded sample.

[0080] Figure 7 According to an embodiment of the present application, a flowchart of actual image changes based on the skeleton method is shown:

[0081] S430: The expanded insulator defect sample data also includes a third expanded sample, see Figure 3 As shown, the third extended sample further includes the following steps:

[0082] Step S431: Acquire multiple natural scene pictures, and group the natural scene pictures into natural scene data;

[0083] Step S432: acquiring second defective insulator image data, and fusing each second insulator defect image in the second defective insulator image data with each natural scene image in the natural scene data to form first simulation data;

[0084] Step S433: acquiring first simulated defective insulator image data, and fusing each first simulated defective insulator image in the first simulated defective insulator image data with each natural scene image in the natural scene data to form second simulated data;

[0085] Step S434: Combine the first simulation data and the second simulation data to form a third expanded sample.

[0086] In this embodiment, the first simulation data acquisition includes the following steps:

[0087] Acquire multiple natural scene images;

[0088] Each second defective insulator image in the second defective insulator image data is fused with each natural scene image to form first simulation data.

[0089] In this embodiment, the third extended sample for simulating and generating a string drop defect includes the following specific steps:

[0090] Obtain the first normal insulator image in the second data; binarize the image of the first normal insulator image in step S10, the image binarization makes the first normal insulator image simple, reduces the amount of data, and highlights the outline of the normal insulator. Thin the edges of the binarized image of the first normal insulator image, and then extract the skeleton image of the normal insulator. Perform restricted region growth; improve the clarity of the skeleton image. Delete the corresponding skeleton points according to the selected roi area; delete the skeleton points corresponding to the selected string drop area in the normal insulator. Remove useless discrete points of the original image according to the skeleton points. Perform post-processing such as edge processing; improve the first simulated defective insulator image after string drop processing. Fuse the generated defective insulator image into the natural background image; the first simulated defective insulator image is fused with each natural background image in the natural background image data to form the data in the third expanded sample.

[0091] S500: Detecting the accuracy of the expanded insulator defect sample data. Detecting the accuracy of the expanded data sample specifically includes the following steps:

[0092] The first expanded sample is used as a defect detection set, and the second expanded sample is used as a test data set;

[0093] Use the defect detection set to detect the accuracy of the test data set;

[0094] When the judgment accuracy is not less than the set threshold;

[0095] The second extended sample is incorporated into the new insulator defect sample.

[0096] S600: Combining the expanded insulator defect sample data with the first data to form a new insulator defect sample. Specifically, combining the first data, the first expanded sample, the second expanded sample and the third expanded sample to form a new insulator defect sample.

[0097] In a possible implementation of the first aspect above, when the judgment accuracy is less than a set threshold;

[0098] The first expanded sample and the second expanded sample are regenerated.

[0099] In some embodiments of the present application, Figure 8 As shown, a system for generating insulator defect samples based on a small amount of data is applied to the method for generating insulator defect samples based on a small amount of data provided in the first aspect, comprising:

[0100] The target detection module 100 is used to divide the existing insulator data into first data containing defective insulators and second data not containing defective insulators;

[0101] A separation module 200, used for separating first defective insulator image data and first background image data in the first data;

[0102] and separating the first normal insulator image data and the second background image data in the second data;

[0103] The expansion module 300 is used to convert the first data into a first expanded sample and the second data into a second expanded sample, and to form a new insulator defect sample by combining the first data, the first expanded sample and the second expanded sample.

[0104] In a possible implementation of the second aspect, a supplementing module 400 is further included, and the supplementing module is used to generate a third expanded sample, and the third expanded sample is used to expand the data volume of the new insulator defect sample.

[0105] In a possible implementation of the second aspect, an evaluation module 500 is further included. The evaluation module uses the first expanded sample as a defect detection set and the second expanded sample as a test data set, and uses the defect detection set to detect the accuracy of the test data set.

[0106] In some embodiments of the present application, an insulator defect sample device based on a small amount of data includes:

[0107] Memory for storing computer programs;

[0108] A processor is used to implement the steps of the method for generating insulator defect samples based on a small amount of data described in the technical solution of the present application when executing a computer program.

[0109] It is understandable that various aspects of the technical solution of the present application can be implemented as a system, method or program product. Therefore, various aspects of the technical solution of the present application can be specifically implemented in the following forms, namely, a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or an implementation combining hardware and software, which can be collectively referred to as "circuit", "module" or "platform" here.

[0110] Fig. 9 According to some embodiments of the present application, a schematic diagram of the structure of an insulator defect sample device based on a small amount of data is shown. Fig. 9 The electronic device 600 implemented according to the implementation in this embodiment is described in detail. Fig. 9 The electronic device 600 shown is merely an example and should not impose any limitations on the functions and scope of use of any embodiment of the technical solution of the present application.

[0111] like Fig. 9As shown, the electronic device 600 is in the form of a general computing device. The components of the electronic device 600 may include but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components (including the storage unit 620 and the processing unit 610), a display unit 640, etc.

[0112] The storage unit stores program codes, which can be executed by the processing unit 610, so that the processing unit 610 executes the implementation steps in the embodiment of the method for generating insulator defect samples with a small amount of data. For example, the processing unit 610 can execute the following steps: Figure 1 Follow the steps shown in .

[0113] The storage unit 620 may include a readable medium in the form of a volatile storage unit, such as a random access unit (RAM) 6201 and / or a cache storage unit 6202 , and may further include a read-only storage unit (ROM) 6203 .

[0114] The storage unit 620 may also include a program / utility 6204 having a set (at least one) of program modules 6205, such program modules 6205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0115] Bus 630 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0116] The electronic device 600 may also communicate with one or more external devices 700 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may communicate with one or more devices that enable a user to interact with the electronic device 600, and / or communicate with any device that enables the electronic device to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed through an input / output (I / O) interface 650. Furthermore, the electronic device 600 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 660. The network adapter 660 may communicate with other modules of the electronic device 600 through the bus 630. It should be understood that although Fig. 9 Not shown, other hardware and / or software modules may be used in conjunction with electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms.

[0117] In some embodiments of the present application, a computer-readable storage medium is provided in the present implementation, on which a computer program is stored. When the computer program is executed by a processor, the relevant steps of the method for generating insulator defect samples based on a small amount of data provided in the above embodiments can be implemented.

[0118] Although this embodiment does not list other specific implementation methods in detail, in some possible implementation methods, various aspects of the technical solution described in the present application can also be implemented in the form of a program product, which includes program code. When the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps described in the image stitching method area of ​​the technical solution of the present application according to the implementation methods in various embodiments of the technical solution of the present application.

[0119] Fig.10 According to some embodiments of the present application, a schematic diagram of the structure of a computer-readable storage medium is shown. Fig.10 As shown, a program product 800 for implementing the above method in an embodiment of the technical solution of the present application is described, which can adopt a portable compact disk read-only memory (CD-ROM) and include program code, and can be run on a terminal device, such as a personal computer. Of course, the program product produced according to this embodiment is not limited to this. In the technical solution of the present application, the readable storage medium can be any tangible medium containing or storing a program, and the program can be used by or in combination with an instruction execution system, an apparatus or a device.

[0120] The program product may use any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0121] Computer readable storage media may include data signals propagated in baseband or as a carrier-area, wherein readable program codes are carried. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The readable storage medium may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, device, or device. The program code contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination thereof.

[0122] The program code for performing the operation of the technical solution of the present application can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and also conventional procedural programming languages ​​such as C language or similar programming languages. The program code can be executed entirely on the user computing device, locally on the user device, as an independent software package, locally on the user computing device, locally on the remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect through the Internet).

[0123] In summary, through the technical solution proposed in this application, it is possible to accurately profile user behavior and obtain portrait groups through group behavior. This technical solution establishes a similar user group for each user, and then realizes automatic classification and acquisition of user portrait groups by setting different aggregation accuracy thresholds. It has the advantages of accurate portraits and adjustable aggregation accuracy. It can obtain various types of user portraits with different accuracy requirements according to different accuracy requirements. It meets the user portrait requirements in product design, content recommendation and other fields, and has promotional value.

[0124] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and one or more of the above embodiments are combined into one embodiment. Those skilled in the art can make various changes or modifications or combinations within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other at will.

Claims

1. A method for generating insulator defect samples based on a small amount of data, characterized in that: The following steps are involved: The existing insulator data is divided into first data containing defective insulators and second data not containing defective insulators; Acquire the first defective insulator image data and the first background image data in the first data by using a target algorithm; Acquire the first normal insulator image data and the second background image data in the second data by using a target algorithm; The expanded insulator defect sample data includes the first expanded sample and the second expanded sample: Performing geometric transformation on each defective insulator image in the first defective insulator image data to form second defective insulator image data, and fusing each second insulator defect image in the second defective insulator image data with each background image in the first background image data to form a first expanded sample; Perform defect simulation on each first normal insulator image in the first normal insulator image data to form first simulated defective insulator image data, and fuse each first simulated defective insulator image in the first simulated defective insulator image data with each second background image in the second background image data to form a second expanded sample; The first data and the expanded insulator defect sample data are combined into a new insulator defect sample.

2. The method for generating insulator defect samples based on a small amount of data according to claim 1, characterized in that: The expanded insulator defect sample data also includes a third expanded sample, and generating the third expanded sample further includes the following steps: Acquire multiple natural scene pictures, and combine the natural scene picture sets into natural scene data; Acquire second defective insulator image data, and fuse each second insulator defect image in the second defective insulator image data with each natural scene image in the natural scene data to form first simulation data; Acquire first simulated defective insulator image data, and fuse each first simulated defective insulator image in the first simulated defective insulator image data with each natural scene image in the natural scene data to form second simulated data; The first simulated data is combined with the second simulated data to form a third expanded sample.

3. The method for generating insulator defect samples based on a small amount of data according to claim 1, characterized in that: Detecting the accuracy of the expanded insulator defect sample data, detecting the accuracy of the expanded insulator defect sample data specifically includes the following steps: Using the first expanded sample as a defect detection set and the second expanded sample as a test data set; Use the defect detection set to detect the accuracy of the test data set; When the judgment accuracy is not less than a set threshold, the expanded insulator defect sample data is incorporated into the new insulator defect sample.

4. The method for generating insulator defect samples based on a small amount of data according to claim 3, characterized in that: When the accuracy rate is determined to be less than a set threshold, the first expanded sample and the second expanded sample are regenerated.

5. The method for generating insulator defect samples based on a small amount of data according to claim 1, characterized in that: The geometric transformation includes flipping, rotation and translation.

6. The method for generating insulator defect samples based on a small amount of data according to claim 1, characterized in that: The defect simulation includes string drop, breakage, shed contamination and self-explosion.

7. An insulator defect sample production system based on a small amount of data, characterized in that: The method for generating insulator defect samples based on a small amount of data as claimed in any one of claims 1 to 6 comprises: A target detection module, used for dividing the existing insulator data into first data containing defective insulators and second data not containing defective insulators; A separation module, used for separating the first defective insulator image data and the first background image data in the first data; and separating the first normal insulator image data and the second background image data in the second data; The expansion module is used to convert the first data into a first expanded sample and the second data into a second expanded sample, and to form a new insulator defect sample by combining the first data, the first expanded sample and the second expanded sample.

8. The insulator defect sample production system based on a small amount of data according to claim 7, characterized in that: The method further comprises a supplementing module, wherein the supplementing module is used to generate a third expanded sample, and the third expanded sample is used to expand the data volume of the new insulator defect sample.

9. The insulator defect sample production system based on a small amount of data according to claim 7, characterized in that: The method further includes an evaluation module, which uses the first expanded sample as a defect detection set, the second expanded sample as a test data set, and uses the defect detection set to detect the accuracy of the test data set.

10. An insulator defect sample generation device based on a small amount of data, characterized in that: include: Memory for storing computer programs; A processor, configured to implement the insulator defect sample generation method based on a small amount of data as described in any one of claims 1 to 6 when executing the computer program.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for generating insulator defect samples based on a small amount of data according to any one of claims 1 to 6 is implemented.

Images