A method and terminal for identifying defects in a transmission line

Abstract

The present invention proposes a method and terminal for identifying defects in power transmission lines, the method comprising: obtaining an image to be identified of the power transmission line; inputting the image to be identified into a preset feature extraction algorithm, setting the area suspected of having defects as a candidate area; classifying and detecting the candidate area based on a trained deep convolutional neural network classifier and an image detection algorithm to distinguish between background areas, normal areas and defective areas. The present invention realizes the rapid and automatic identification of defects in power transmission lines by means of automatic image recognition, thereby improving efficiency and reducing costs.

Description

Technical Field

[0001] The present invention relates to the technical field of power transmission line defect detection, and in particular to a power transmission line defect identification method and terminal. Background Art

[0002] Transmission lines provide a steady supply of electricity for people's work and life. Therefore, transmission lines are widely used in modern society. However, with the transmission lines themselves and with their use in different environments, especially harsh environments, some defects will appear in the transmission lines. These defects will affect the normal power supply. Therefore, it is necessary to identify the defects of the transmission lines in time.

[0003] At present, in the identification of defects in power transmission lines, the more conventional method is to use manual methods. Specifically, experienced staff are dispatched to inspect the power transmission lines on the spot with the naked eye, or to view pictures of the power transmission lines on a display to manually detect whether there are defects. However, this method is inefficient and requires a lot of manpower. At present, the number of transmission lines is huge, and the manual method cannot effectively cope with the identification needs of such a large number of transmission lines. At present, there are also methods of identification through drone photography, such as the Chinese patent with publication number CN109544501A and publication date 2019.3.29 - A method for detecting defects in power transmission equipment based on multi-source image feature matching of drones. This patent uses drones to take pictures and directly analyze and identify the pictures. Although no manpower is required, the recognition accuracy is not high and the efficiency is low. For this reason, a better solution is currently needed to identify power transmission lines. Summary of the invention

[0004] The present invention proposes a transmission line defect recognition method and terminal, which realizes rapid and automatic recognition of defects in the transmission line in an automated image recognition manner, thereby improving efficiency and reducing costs.

[0005] Specifically, the present invention proposes the following specific technical solutions:

[0006] A method for identifying defects in a power transmission line, comprising:

[0007] Acquire an image of a transmission line to be identified;

[0008] The image to be identified is input into a preset feature extraction algorithm, and the area suspected of having defects is set as a candidate area;

[0009] Based on the trained deep convolutional neural network classifier and image detection algorithm, the candidate areas are classified and detected to distinguish between background areas, normal areas and defect areas.

[0010] In a specific embodiment, the preset feature extraction algorithm is obtained through artificial experience knowledge and based on a color space decomposition method.

[0011] In a specific embodiment, the area suspected of having defects is any combination of one or more of the following: an area with conductor strand breakage ablation features, an area with insulator porcelain bottle lightning ablation features, an area with damaged geometric features, an area with specific color features, an area with specific texture features, and an area with specific shape features.

[0012] In a specific embodiment, it also includes:

[0013] Acquire a sample image of the power transmission line; wherein the sample image includes a forward sample image in which defects are identified and a reverse sample image in which defects are determined not to exist;

[0014] The forward sample image and the reverse sample image are trained in a deep convolutional neural network to obtain a trained deep convolutional neural network classifier.

[0015] In a specific embodiment, the number of positive sample images is greater than the number of negative sample images.

[0016] In a specific embodiment, the ratio of the number of positive sample images to the number of negative sample images is 4:1.

[0017] In a specific embodiment, the specific process of classifying and detecting the candidate areas based on the trained deep convolutional neural network classifier and the image detection algorithm to distinguish the background area, the normal area and the defect area includes:

[0018] The candidate region is processed by a trained deep convolutional neural network classifier, and feature data of the penultimate layer of the deep neural network in the deep convolutional neural network classifier is extracted;

[0019] The feature data is classified and detected through the image detection algorithm to distinguish the background area, normal area and defect area.

[0020] In a specific embodiment, the image detection algorithm includes: an image detection method based on a BP neural network or an image detection method based on a SVM.

[0021] In a specific embodiment, the SVM-based image detection method detects defects in a power transmission line, specifically including:

[0022] Respectively intercepting a plurality of broken strand images and a plurality of non-broken strand images of a transmission line of a preset length;

[0023] After extracting features from the broken strand images and the non-broken strand images, the SVM algorithm is used for training to obtain the SVM recognition model;

[0024] The image of the transmission line to be identified is pre-processed and then input into the SVM recognition model to compare and determine the image area of ​​the transmission line of the preset length in the identification image;

[0025] If the image area is smaller than the image area of ​​the preset length of a normal transmission line, it is determined that the transmission line corresponding to the image to be identified is a broken transmission line; wherein, the image area of ​​the preset length of a normal transmission line is obtained based on feature extraction of the non-broken image and training using the SVM algorithm.

[0026] The present invention also proposes a transmission line defect identification terminal, which includes an acquisition module, a candidate processing module, and an identification and classification module, wherein:

[0027] An acquisition module, used for acquiring an image to be identified of a transmission line;

[0028] A candidate processing module, used for inputting the image to be identified into a preset feature extraction algorithm, and setting the area suspected of having defects as a candidate area;

[0029] The recognition and classification module uses the trained deep convolutional neural network classifier and image detection algorithm to classify and detect candidate areas to distinguish between background areas, normal areas and defect areas.

[0030] The present invention also provides a computer storage medium, which stores a program for executing the above method.

[0031] The beneficial effects of the present invention are:

[0032] The present invention proposes a method and terminal for identifying defects in a power transmission line. The present invention first processes an image to be identified to determine areas suspected of having defects, and excludes areas that do not need to be processed. Then, only the determined areas need to be classified and detected based on a deep convolutional neural network classifier and an image detection algorithm, thereby realizing automatic recognition of images of power transmission lines. As long as the image to be identified is input, batch and rapid recognition can be performed. Compared with manual recognition, rapid and automatic recognition of defects in power transmission lines is realized, thereby improving efficiency and reducing costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a schematic diagram of the flow chart of the power transmission line defect identification method of the present invention;

[0034] Figure 2 It is a structural schematic diagram of the mobile phone terminal of the present invention;

[0035] Figure 3 It is a schematic diagram of the functional framework structure of the power transmission line defect identification terminal of the present invention;

[0036] Figure 4 It is a schematic diagram of the structure of the computer storage medium of the present invention. DETAILED DESCRIPTION

[0037] In the following, various embodiments of the present disclosure will be described more fully. The present disclosure may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of the present disclosure to the specific embodiments disclosed herein, but rather the present disclosure should be understood to cover all adjustments, equivalents and / or alternatives that fall within the spirit and scope of the various embodiments of the present disclosure.

[0038] The terms used in the various embodiments of the present disclosure are only used for the purpose of describing specific embodiments and are not intended to limit the various embodiments of the present disclosure. As used herein, the singular form is intended to also include the plural form, unless the context clearly indicates otherwise. Unless otherwise defined, all terms (including technical terms and scientific terms) used herein have the same meaning as the meanings commonly understood by ordinary technicians in the field to which the various embodiments of the present disclosure belong. The terms (such as the terms defined in the generally used dictionary) will be interpreted as having the same meaning as the contextual meaning in the relevant technical field and will not be interpreted as having an idealized meaning or an overly formal meaning, unless clearly defined in the various embodiments of the present disclosure.

[0039] Example 1

[0040] A multi-load data fusion method for transmission line defect identification, such as Figure 1 As shown, including:

[0041] Step 101: obtaining an image of a power transmission line to be identified;

[0042] Specifically, the image to be identified may be obtained by photographing a power transmission line.

[0043] Step 102: input the image to be identified into a preset feature extraction algorithm, and set the area suspected of having defects as a candidate area;

[0044] Specifically, the preset feature extraction algorithm is obtained through artificial experience knowledge and based on the color space decomposition method, and the suspected defective area is any combination of one or more of the following: an area with ablation features of loose wire strands, an area with ablation features of insulator porcelain bottles due to lightning strikes, an area with damaged geometric features, an area with specific color features, an area with specific texture features, and an area with specific shape features. Through step 102, the candidate area can be determined to reduce the subsequent processing pressure and improve the processing efficiency.

[0045] Step 103: Classify and detect the candidate areas based on the trained deep convolutional neural network classifier and image detection algorithm to distinguish between background areas, normal areas and defective areas.

[0046] In this embodiment, a sample image of the power transmission line is obtained; wherein the sample image includes a forward sample image in which defects are marked and a reverse sample image in which defects are determined not to exist;

[0047] By training the forward sample images and the reverse sample images in the deep convolutional neural network, a trained deep convolutional neural network classifier is obtained.

[0048] In this way, the trained deep convolutional neural network classifier can well identify the defects of the transmission line;

[0049] In order to further improve the recognition accuracy and improve the training efficiency, the number of positive sample images is greater than the number of negative sample images. Preferably, the ratio of the number of positive sample images to the number of negative sample images is 4:1. Practice has proved that the ratio of the number of positive sample images to the number of negative sample images is 4:1, which can comprehensively improve the training efficiency and the number of samples, and can achieve a good training effect.

[0050] In this embodiment, step 103 specifically includes:

[0051] The candidate region is processed by a trained deep convolutional neural network classifier, and feature data of the penultimate layer of the deep neural network in the deep convolutional neural network classifier is extracted;

[0052] The feature data is classified and detected through the image detection algorithm to distinguish the background area, normal area and defect area.

[0053] In this embodiment, the image detection algorithm includes: an image detection method based on a BP neural network or an image detection method based on an SVM. For example, a support vector machine (SVM) is used to map feature vectors to a high-dimensional space for classification.

[0054] In this embodiment, a deep neural network is used to extract target features in the candidate area in an all-round way. In most cases, the integrity and effectiveness of the features obtained are better than those extracted manually. However, considering that the last layer of the deep neural network is usually in the form of a fully connected layer of a traditional neural network, classification is performed through Softmax, which is easy to generate or fit when the number of training samples is small.

[0055] Taking this into consideration, in this solution, the feature data of the penultimate layer of the deep neural network is extracted, and then the conventional image detection algorithm is used to classify and detect the feature data to distinguish between background areas, normal areas and defective areas (in this regard, it should be noted that the extraction of the feature data of the penultimate layer of the deep neural network is one of the important technical innovations of this embodiment, which selects and optimizes the source of the feature data, so it has important technical significance for subsequent identification). This can fully utilize the advantages of traditional methods in feature classification, and combine the advantages of deep neural networks in image feature extraction to realize the automatic identification of typical power transmission defects that integrates traditional means and modern technologies, and improve the accuracy of typical power transmission defect identification.

[0056] Therefore, the present invention studies the differentiated characteristics of different processing stages such as image segmentation, feature extraction, and defect recognition of typical wire defects based on traditional methods (such as BP neural network, support vector machine SVM, etc.) and deep neural network method (convolutional neural network CNN), and studies a multi-technology fusion recognition scheme for wire defects such as broken strands, loose strands, and ablation, giving full play to the advantages of traditional methods and deep neural network methods in each stage of image recognition, and improving the recognition accuracy of different defect types.

[0057] The following is a further description of the SVM algorithm used in this embodiment to realize defect recognition of typical defect images of conductors: The SVM-based image detection method detects defects in transmission lines, specifically including:

[0058] Respectively intercepting a plurality of broken strand images and a plurality of non-broken strand images of a transmission line of a preset length;

[0059] After extracting features from the broken strand image and the non-broken strand image, training is performed using an SVM algorithm to obtain an SVM recognition model;

[0060] The image of the transmission line to be identified is pre-processed and then input into the SVM recognition model to compare and determine the image area of ​​the transmission line of the preset length in the identification image;

[0061] If the image area is smaller than the image area of ​​the preset length of a normal transmission line, it is determined that the transmission line corresponding to the image to be identified is a broken transmission line; wherein, the image area of ​​the preset length of a normal transmission line is obtained based on feature extraction of the non-broken image and training using the SVM algorithm.

[0062] The defect area can be obtained through the above step 103; the specific process of identifying the broken strands of the wire segment in the present embodiment of step 103 is exemplified as follows:

[0063] Research has found that the image area of ​​a wire with a broken strand will become smaller for a predetermined length; however, the normal length-width area of ​​a wire with a preset length is a fixed value, and there is a minimum domain threshold. This minimum domain threshold can be obtained through a large amount of data training; the SVM algorithm can then be used to obtain the wire image area with broken strands. The specific operation method is as follows:

[0064] First, construct the SVM algorithm recognition training set and test set and perform training. Take m images of broken strands with a preset length of A×A pixels and n images of non-broken strands with a preset length to form the test set. Then, extract features from the intercepted images. Then, use the SVM algorithm for training to determine the SVM recognition model.

[0065] Then the image to be identified is preprocessed (the image to be identified is input into the SVM recognition model as a test set), the preprocessed image is segmented, and the rectangular area to be identified is determined to accurately determine the position of the wire;

[0066] Finally, the broken strand image and image background of the wire image of the preset length are separated; finally, the SVM algorithm recognition model is used to identify the broken strand anomaly of the preset length. If it is found that the current wire image successfully matches the image in the standard data set, it can be judged as a broken strand wire (that is, the current wire image that successfully matches a series of typical defect images below the minimum domain threshold can be judged as a broken strand wire).

[0067] It should be specially noted that the preprocessing of the image to be identified includes: using the background difference method to remove the background of the broken strand image to obtain a primary filtered image; performing threshold segmentation on the primary filtered image to obtain a secondary filtered image; counting the area size of each connected region of the pixels in the secondary filtered image, estimating the area of ​​the wire region of a preset length, and extracting the small area (i.e. extracting a series of images of typical defects below the minimum domain threshold and using them as the final standard data set of typical defective wires).

[0068] Example 2

[0069] The present invention also provides a transmission line defect identification terminal, such as Figure 2For the sake of convenience, only the parts related to the embodiments of the present invention are shown. For specific technical details not disclosed, please refer to the method part of the embodiments of the present invention. The terminal can be any terminal device including a mobile phone, a tablet computer, a PDA (Personal Digital Assistant), a POS (Point of Sales), a car computer, etc.

[0070] Take the case where the terminal is a mobile phone as an example:

[0071] Figure 2 FIG. 1 is a block diagram showing a partial structure of a mobile phone related to a terminal provided in an embodiment of the present invention. Figure 2 The mobile phone includes: Radio Frequency (RF) circuit, memory, input unit, display unit, sensor, audio circuit, WiFi module, processor and other components. Those skilled in the art can understand that Figure 2 The mobile phone structure shown in the figure does not constitute a limitation on the mobile phone, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0072] Combine the following Figure 2 A detailed introduction to the various components of the mobile phone:

[0073] RF circuits can be used to receive and send signals during the process of sending and receiving information or making calls. In particular, after receiving the downlink information of the base station, it is processed by the processor; in addition, the designed uplink data is sent to the base station. Generally, the RF circuit includes but is not limited to an antenna, at least one amplifier, a transceiver, a coupler, a low noise amplifier (Low Noise Amplifier, LNA), a duplexer, etc. In addition, the RF circuit can also communicate with the network and other devices through wireless communication. The above-mentioned wireless communication can use any communication standard or protocol, including but not limited to the Global System of Mobilecommunication (GSM), General Packet Radio Service (GPRS), Code Division Multiple Access (CDMA), Wideband Code Division Multiple Access (WCDMA), Long Term Evolution (LTE), email, Short Messaging Service (SMS), etc.

[0074] The memory can be used to store software programs and modules. The processor executes various functional applications and data processing of the mobile phone by running the software programs and modules stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0075] The input unit can be used to receive input digital or character information, and to generate key signal input related to the user settings and function control of the mobile phone. Specifically, the input unit may include a touch panel and other input devices. The touch panel, also known as a touch screen, can collect the user's touch operation on or near it (such as the user's operation on or near the touch panel using any suitable object or accessory such as a finger, stylus, etc.), and drive the corresponding connection terminal according to a pre-set program. Optionally, the touch panel may include two parts: a touch detection terminal and a touch controller. Among them, the touch detection terminal detects the user's touch orientation, detects the signal brought by the touch operation, and transmits the signal to the touch controller; the touch controller receives the touch information from the touch detection terminal, converts it into contact coordinates, and then sends it to the processor, and can receive and execute commands sent by the processor. In addition, the touch panel can be implemented using multiple types such as resistive, capacitive, infrared, and surface acoustic waves. In addition to the touch panel, the input unit may also include other input devices. Specifically, other input devices may include, but are not limited to, one or more of a physical keyboard, function keys (such as volume control keys, switch keys, etc.), a trackball, a mouse, a joystick, etc.

[0076] The display unit can be used to display information input by the user or information provided to the user and various menus of the mobile phone. The display unit may include a display panel. Optionally, the display panel may be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), etc. Further, the touch panel may cover the display panel. When the touch panel detects a touch operation on or near it, it is transmitted to the processor to determine the type of touch event, and then the processor provides a corresponding visual output on the display panel according to the type of touch event. Although in Figure 2In the embodiment, the touch panel and the display panel are used as two independent components to realize the input and output functions of the mobile phone, but in some embodiments, the touch panel and the display panel can be integrated to realize the input and output functions of the mobile phone.

[0077] The mobile phone may also include at least one sensor, such as a light sensor, a motion sensor, and other sensors. Specifically, the light sensor may include an ambient light sensor and a proximity sensor, wherein the ambient light sensor may adjust the brightness of the display panel according to the brightness of the ambient light, and the proximity sensor may turn off the display panel and / or backlight when the mobile phone is moved to the ear. As a type of motion sensor, the accelerometer sensor can detect the magnitude of acceleration in all directions (generally three axes), and can detect the magnitude and direction of gravity when stationary. It can be used for applications that identify the posture of the mobile phone (such as horizontal and vertical screen switching, related games, magnetometer posture calibration), vibration recognition related functions (such as pedometer, tapping), etc.; as for other sensors that can be configured in the mobile phone, such as gyroscopes, barometers, hygrometers, thermometers, infrared sensors, etc., they will not be repeated here.

[0078] The audio circuit, speaker, and microphone can provide an audio interface between the user and the mobile phone. The audio circuit can convert the received audio data into an electrical signal and transmit it to the speaker, which converts it into a sound signal for output; on the other hand, the microphone converts the collected sound signal into an electrical signal, which is received by the audio circuit and converted into audio data. The audio data is then processed by the output processor and sent to, for example, another mobile phone through the RF circuit, or the audio data is output to the memory for further processing.

[0079] WiFi is a short-range wireless transmission technology. Mobile phones can help users send and receive emails, browse web pages, and access streaming media through WiFi modules. It provides users with wireless broadband Internet access. Figure 2 A WiFi module is shown, but it is understandable that it is not an essential component of the mobile phone and can be omitted as needed without changing the essence of the invention.

[0080] The processor is the control center of the mobile phone. It uses various interfaces and lines to connect the various parts of the entire mobile phone. It executes various functions of the mobile phone and processes data by running or executing software programs and / or modules stored in the memory, and calling data stored in the memory, so as to monitor the mobile phone as a whole. Optionally, the processor may include one or more processing units; preferably, the processor may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface and application programs, etc., and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor.

[0081] The mobile phone also includes a power source (such as a battery) for supplying power to various components. Preferably, the power source can be logically connected to the processor through a power management system, so that the power management system can manage charging, discharging, power consumption and other functions.

[0082] Although not shown, the mobile phone may also include a camera, a Bluetooth module, etc., which will not be described in detail here.

[0083] In the embodiment of the present invention, the processor included in the mobile terminal also has the following functions: Figure 3 As shown, the terminal includes an acquisition module, a candidate processing module, and a recognition and classification module; the acquisition module is used to acquire the image to be identified of the transmission line; the candidate processing module is used to input the image to be identified into a preset feature extraction algorithm, and set the area suspected of having defects as a candidate area; the recognition and classification module is used to classify and detect the candidate area using the trained deep convolutional neural network classifier and the image detection algorithm to distinguish between the background area, the normal area and the defective area.

[0084] In a specific embodiment, the preset feature extraction algorithm is obtained by artificial experience knowledge and based on the color space decomposition method. In a specific embodiment, the area suspected of having defects is any combination of one or more of the following: an area with ablation features of loose wire strands, an area with ablation features of insulator porcelain bottles due to lightning strikes, an area with damaged geometric features, an area with specific color features, an area with specific texture features, and an area with specific shape features.

[0085] In a specific embodiment, the method further includes: obtaining sample images of the power transmission line; wherein the sample images include positive sample images with defects and negative sample images with defects; and obtaining a trained deep convolutional neural network classifier by training the positive sample images and the negative sample images in a deep convolutional neural network. The number of positive sample images is greater than the number of negative sample images, and the ratio of the number of positive sample images to the number of negative sample images is 4:1.

[0086] In a specific embodiment, the candidate area is processed by a trained deep convolutional neural network classifier, and the feature data of the penultimate layer of the deep neural network in the deep convolutional neural network classifier is extracted; the feature data is classified and detected by an image detection algorithm to distinguish between background areas, normal areas and defective areas. The image detection algorithm includes: an image detection method based on a BP neural network or an image detection method based on a SVM.

[0087] The image detection method based on SVM detects defects in power transmission lines, including:

[0088] A plurality of broken strand images and a plurality of non-broken strand images of a transmission line of a preset length are respectively intercepted; after feature extraction of the broken strand images and the non-broken strand images, an SVM algorithm is used for training to obtain an SVM recognition model; the image to be identified of the transmission line is pre-processed and then input into the SVM recognition model to compare and determine the image region area of ​​the transmission line of the preset length in the identification image; if the image region area is smaller than the image region area of ​​the preset length of a normal transmission line, it is determined that the transmission line corresponding to the image to be identified is a broken strand transmission line; wherein the image region area of ​​the preset length of a normal transmission line is obtained based on feature extraction of the non-broken strand image and training with the SVM algorithm.

[0089] Example 3

[0090] The present invention also provides a computer storage medium, which stores a program for executing the method in embodiment 1.

[0091] Specifically, Figure 4 A computing device that can implement the processing method for adaptively identifying harassing calls according to the present invention is shown. The computing device traditionally includes a processor and a computer program product or a computer-readable medium in the form of a storage device. The storage device may be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read-only memory), an EPROM, a hard disk, or a ROM. The storage device has a storage space for storing program codes for executing any method steps in the above method. For example, the storage space for storing program codes may include individual program codes for implementing various steps in the above method, respectively. These program codes can be read from or written to one or more computer program products. These computer program products include program code carriers such as a hard disk, a compact disk (CD), a memory card, or a floppy disk. Such a computer program product is typically a portable or fixed storage unit. The storage unit may have a storage space with Figure 4 The storage device in the computing device of the computer may be similarly arranged as a storage segment, storage space, etc. The program code may be compressed, for example, in an appropriate form. Generally, the storage unit includes computer readable codes for executing the steps of the method according to the present invention, that is, codes that can be read by a processor such as the computer, and when these codes are executed by the computing device, the computing device is caused to execute the various steps in the method described in embodiment 1.

[0092] Those skilled in the art will appreciate that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily necessary for the implementation of the present invention. Those skilled in the art will appreciate that the modules in the terminals in the implementation scenario can be distributed in the terminals of the implementation scenario according to the description of the implementation scenario, or can be changed accordingly and located in one or more terminals different from the present implementation scenario. The modules of the above-mentioned implementation scenarios can be combined into one module, or can be further divided into multiple sub-modules. The above-mentioned serial numbers of the present invention are only for description and do not represent the advantages and disadvantages of the implementation scenarios. Disclosed above are only a few specific implementation scenarios of the present invention, but the present invention is not limited thereto, and any changes that can be thought of by those skilled in the art should fall within the scope of protection of the present invention.

Claims

1. A method for identifying defects in a power transmission line, characterized in that: include: Acquire an image of a transmission line to be identified; The image to be identified is input into a preset feature extraction algorithm, and the area suspected of having defects is set as a candidate area; Classify and detect candidate areas based on the trained deep convolutional neural network classifier and image detection algorithm to distinguish between background areas, normal areas and defect areas; The specific process of classifying and detecting candidate areas based on the trained deep convolutional neural network classifier and image detection algorithm to distinguish between background areas, normal areas and defect areas includes: The candidate region is processed by a trained deep convolutional neural network classifier, and feature data of the penultimate layer of the deep neural network in the deep convolutional neural network classifier is extracted; Classify and detect feature data through image detection algorithms to distinguish between background areas, normal areas and defect areas; Image detection algorithms include: SVM-based image detection method; The image detection method based on SVM detects defects in power transmission lines, including: Respectively intercepting a plurality of broken strand images and a plurality of non-broken strand images of a transmission line of a preset length; After extracting features from the broken strand images and the non-broken strand images, the SVM algorithm is used for training to obtain the SVM recognition model; The image of the transmission line to be identified is pre-processed and then input into the SVM recognition model to compare and determine the image area of ​​the transmission line of the preset length in the identification image; If the image area is smaller than the image area of ​​the preset length of a normal transmission line, it is determined that the transmission line corresponding to the image to be identified is a broken transmission line; wherein, the image area of ​​the preset length of a normal transmission line is obtained based on feature extraction of the non-broken image and training using the SVM algorithm.

2. The defect identification method according to claim 1, characterized in that: The preset feature extraction algorithm is obtained through artificial experience knowledge and based on the color space decomposition method.

3. The defect identification method according to claim 1, characterized in that: The area suspected of having defects is any combination of one or more of the following: an area with conductor strand breakage ablation features, an area with insulator porcelain bottle lightning ablation features, an area with damaged geometric features, an area with specific color features, an area with specific texture features, and an area with specific shape features.

4. The defect identification method according to claim 1, characterized in that: Also includes: Acquire a sample image of the power transmission line; wherein the sample image includes a forward sample image in which defects are identified and a reverse sample image in which defects are determined not to exist; By training the forward sample images and the reverse sample images in the deep convolutional neural network, a trained deep convolutional neural network classifier is obtained.

5. The defect identification method according to claim 4, characterized in that: The number of positive sample images is greater than that of negative sample images.

6. The defect identification method according to claim 4 or 5, characterized in that: The ratio of the number of positive sample images to the number of negative sample images is 4:

1.

7. A power transmission line defect identification terminal, used to execute the defect identification method according to any one of claims 1 to 6, characterized in that: The terminal includes an acquisition module, a candidate processing module, and an identification and classification module, wherein: An acquisition module, used for acquiring an image to be identified of a transmission line; A candidate processing module, used for inputting the image to be identified into a preset feature extraction algorithm, and setting the area suspected of having defects as a candidate area; The recognition and classification module uses the trained deep convolutional neural network classifier and image detection algorithm to classify and detect candidate areas to distinguish between background areas, normal areas and defect areas.

Images