Wildfire identification system, method and related device based on unmanned aerial vehicle inspection

Abstract

The embodiment of the application discloses a mountain fire identification system and method based on unmanned aerial vehicle inspection and related devices, the mountain fire identification system comprises: an unmanned aerial vehicle data acquisition system, a deep learning analysis system and a terminal visualization system; the unmanned aerial vehicle data acquisition system is used for collecting image data of a power transmission line; the deep learning analysis system is used for identifying mountain fires in the image data by using a pre-trained deep learning model, and determining a mountain fire identification result of the power transmission line for reflecting whether the power transmission line has a mountain fire; and the terminal visualization system is used for generating a visual prompt based on the mountain fire identification result. Through the above mountain fire identification system combining a deep learning algorithm and unmanned aerial vehicle data acquisition, potential dangerous situations such as mountain fires can be automatically identified, and an alarm can be sent in time, thereby significantly improving the safety of the power transmission line and the efficiency of the inspection work, and providing an efficient and intelligent mountain fire automatic identification and alarm system for the power transmission line.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of unmanned aerial vehicle inspection, and particularly relates to a mountain fire identification system and method based on unmanned aerial vehicle inspection and a related device. BACKGROUND

[0002] Mountain fires have always been one of the main reasons for the tripping of high-voltage transmission lines. When a mountain fire occurs in the transmission line corridor, the reclosing success rate of the protection device is very low, and the reasons are multifaceted. First, the reclosing time is generally on the order of seconds or even milliseconds, while the duration of a mountain fire is usually several minutes or even hours, far exceeding the time range of reclosing. Therefore, even if the reclosing is successful for a short time, the continuous spread of the fire may lead to tripping again, affecting the stability of the power grid.

[0003] Secondly, the high temperature and dense smoke generated by mountain fires not only can directly damage transmission facilities such as conductors and towers, but also can cause the performance of insulating materials to decline, increasing the risk of short-circuit tripping. In addition, mountain fires can also change the conditions of the surrounding environment, such as air composition and humidity, which may affect the normal operation of electrical equipment and increase the failure rate. Furthermore, the occurrence of mountain fires is often accompanied by extreme weather conditions such as drought and strong winds, which are not conducive to the safe operation of power grid facilities.

[0004] In the current high-voltage transmission line inspection, the photos of the transmission line are usually obtained by manual inspection, fixed camera monitoring, and unmanned aerial vehicle shooting, and then the presence or absence of mountain fires that endanger the safe and stable operation of the transmission line is determined by manual visual recognition. However, the manual inspection method is labor-intensive and requires real-time photographing by the staff, and the fixed camera shooting requires a large investment and high reliability of the device. The existing unmanned aerial vehicle inspection system relies on manual analysis of video data, which is inefficient and easily affected by subjective factors. Mountain fire detection of high-voltage transmission lines is an important sensing means for obtaining information about the transmission line on site, which not only provides a basis for power grid operation mode adjustment, emergency plan development, line operation and maintenance state evaluation, but also provides support for government departments to start fire fighting decisions.

[0005] Currently, the photo information of the power transmission line is obtained through artificial patrol, fixed camera monitoring and unmanned aerial vehicle shooting, and is reported through telephone and WeChat group. At the same time, the data processing personnel confirm the position of the forest fire monitoring through the patrol position of the forest fire observer, the position of the fixed camera and the shooting direction of the unmanned aerial vehicle. The data processing personnel combine the obtained photo information of the power transmission line with the detection position to form an excel summary table of forest fire observation, and then manually input it into the forest fire management platform of the power grid. Professional personnel then review the uploaded and input information to determine whether there is a forest fire and whether the forest fire will affect the safe and stable operation of the power transmission line, and then make a decision. The above traditional forest fire monitoring method has the following defects:

[0006] 1) The traditional inspection method cannot realize full-time and full-range monitoring, and there is a monitoring blind area. The traditional power transmission line inspection method, such as artificial patrol or fixed camera monitoring, has significant limitations. Artificial patrol not only consumes time and effort, but also cannot cover all key areas due to geographical and environmental conditions, especially in remote or difficult-to-reach places. Although fixed cameras can provide real-time monitoring, their field of view is limited by the installation location and cannot adjust the viewing angle, which can easily form a monitoring blind area. In addition, fixed cameras have insufficient tracking ability for moving targets, and it is difficult to capture the dynamic changes of the fire source in real time once an emergency such as a forest fire occurs. For power transmission lines that span a wide area, full-time and full-range monitoring is crucial to ensure their safe operation. The traditional inspection method cannot meet this demand, so there is a significant safety hazard.

[0007] 2) Artificial analysis of video data is time-consuming and labor-intensive, and it is difficult to ensure accuracy and consistency. In the traditional monitoring system, analyzing video data usually relies on manual operation. This process not only consumes a lot of time and human resources, but also due to human factors, the accuracy and consistency of the results are difficult to guarantee. Human fatigue, distraction or subjective judgment can all affect the quality of analysis. Especially in emergency situations such as forest fires, quickly and accurately identifying the location of the fire source and the development of the fire is crucial for a timely response. The manual analysis method often cannot process a large amount of video data in a short period of time and extract key information from it, which may lead to the spread of the fire and cause greater losses.

[0008] 3) Existing UAV inspection systems lack automated wildfire detection capabilities and cannot respond to wildfire threats in a timely manner. Although UAV inspection systems have improved inspection efficiency and coverage to some extent, most existing UAV systems rely on manual remote control or post-data analysis. These systems lack automated wildfire detection capabilities, that is, the function of real-time detection, identification and reporting of wildfire events during flight. Wildfires occur suddenly and spread rapidly, requiring inspection systems to respond quickly. Manual control of UAVs or post-data analysis methods cannot meet the demand for rapid wildfire identification and timely response. Therefore, UAV inspection systems lacking automated identification capabilities have obvious shortcomings in handling emergencies such as wildfires.

[0009] 4) Data processing and analysis speed is insufficient to meet real-time monitoring needs. Data is the core of modern monitoring systems, especially when dealing with emergencies such as wildfires, the speed of data processing and analysis is crucial. However, existing UAV inspection systems often have bottlenecks in data processing and analysis. Due to inefficient algorithms or limited computing resources, these systems cannot process and analyze data immediately after collection, resulting in delayed response. Real-time monitoring requires systems to quickly process large amounts of data, identify abnormal situations in a timely manner, and respond accordingly. If the speed of data processing and analysis cannot keep up with the speed of data collection, the system cannot achieve true real-time monitoring, missing the best opportunity to prevent and control wildfires.

[0010] Therefore, the traditional wildfire monitoring method is inefficient, and the wildfire identification accuracy is also low, and there is an urgent need for a wildfire monitoring method that can improve efficiency and accuracy. SUMMARY

[0011] The main purpose of the present application is to provide a wildfire identification system, method and related device based on UAV inspection, which can solve the problem of low efficiency and low accuracy of wildfire identification in the prior art.

[0012] To achieve the above-mentioned purpose, the first aspect of the present application provides a wildfire identification system based on UAV inspection, which comprises: a UAV data acquisition system, a deep learning analysis system and a terminal visualization system, and the UAV data acquisition system, the deep learning analysis system and the terminal visualization system are communicatively connected;

[0013] The UAV data acquisition system is used for acquiring image data of the power transmission line, and sending the image data to the deep learning analysis system;

[0014] The deep learning analysis system is configured to receive the image data, perform a wildfire identification on the image data by using a pre-trained deep learning model, determine a wildfire identification result of the power transmission line, and send the wildfire identification result to the terminal visualization system, wherein the wildfire identification result is used to reflect whether there is a wildfire on the power transmission line, and the deep learning model is pre-trained based on historical wildfire image data of the power transmission line.

[0015] The terminal visualization system is configured to receive the wildfire identification result and generate a visual prompt based on the wildfire identification result.

[0016] In an implementable manner, the unmanned aerial vehicle data acquisition system comprises at least a high-voltage power transmission line account module, a sensor camera control module, a data acquisition and storage unit module, and a data communication upload module, wherein the high-voltage power transmission line account module, the sensor camera control module, the data acquisition and storage unit module, and the data communication upload module are communicatively connected.

[0017] The high-voltage power transmission line account module is configured to store account information of a high-voltage power transmission line and send the account information to the sensor camera control module.

[0018] The sensor camera control module is configured to receive the account information, perform data acquisition planning processing based on the account information to obtain data acquisition instructions, perform image acquisition processing on the power transmission line based on the data acquisition instructions to obtain the image data, and send the image data to the data acquisition and storage unit module.

[0019] The data acquisition and storage unit module is configured to receive and store the image data and send the image data to the data communication upload module.

[0020] The data communication upload module is configured to receive the image data and upload the image data to the deep learning analysis system.

[0021] In an implementable manner, the unmanned aerial vehicle data acquisition system further comprises a data analysis and processing module, wherein an input end of the data analysis and processing module is communicatively connected to an output end of the data acquisition and storage unit module, and the output end of the data acquisition and storage unit module is communicatively connected to an input end of the data communication upload module.

[0022] The data acquisition and storage unit module is further configured to send the image data to the data analysis and processing module.

[0023] The data analysis processing module is configured to receive the image data, perform denoising processing on the image data to obtain denoised image data, and send the denoised image data to the data communication uploading module.

[0024] The data communication uploading module is configured to receive the denoised image data and upload the denoised image data to the deep learning analysis system.

[0025] In an implementable manner, the deep learning analysis system at least includes a data preprocessing module, a data labeling processing module, an image cutting processing module, a neural network feature extraction module, a model training module, a regression prediction module, and an output analysis result module, wherein the data preprocessing module, the data labeling processing module, the image cutting processing module, the neural network feature extraction module, the model training module, the regression prediction module, and the output analysis result module are communicatively connected.

[0026] The data preprocessing module is configured to receive the historical forest fire image data and the image data, preprocess the historical forest fire image data to obtain preprocessed historical forest fire image data.

[0027] The data labeling processing module is configured to label the preprocessed historical forest fire image data with training labels to obtain historical forest fire image data labeled with training labels.

[0028] The image cutting processing module is configured to cut the historical forest fire image data labeled with training labels into image blocks according to a preset image segmentation rule.

[0029] The neural network feature extraction module is configured to extract features of the image blocks using a neural network model to obtain image features of the image blocks.

[0030] The model training module is configured to train a deep learning model for forest fire recognition using the image features and the training labels to obtain a trained deep learning model.

[0031] The regression prediction module is configured to perform regression prediction using the trained deep learning model and the image data to determine a forest fire recognition result of the power transmission line.

[0032] The output analysis result module is configured to send the forest fire recognition result to the terminal visual system.

[0033] In an implementable manner, the terminal visual system at least includes a short message alarm pushing module and a visual display module, and the short message alarm pushing module and the visual display module are communicatively connected.

[0034] The short message alarm pushing module is configured to receive the wildfire identification result, and output a short message alarm information to the visual display module according to the wildfire identification result and a current operation state of the power transmission line.

[0035] The visual display module is configured to receive the short message alarm information, and output a visual prompt based on the short message alarm information.

[0036] In a feasible implementation manner, the terminal visual system further comprises an auxiliary decision display module and a feedback loop module, and the auxiliary decision display module and the feedback loop module are communicatively connected;

[0037] The auxiliary decision display module is configured to receive an artificial auxiliary decision result, and send the artificial auxiliary decision result to the feedback loop module, where the artificial auxiliary decision result is used to indicate a confirmation result of an artificial auxiliary decision on accuracy of the wildfire identification result.

[0038] The feedback loop module is configured to receive the artificial auxiliary decision result, and optimize training of the trained deep learning model by taking the artificial auxiliary decision result as feedback data, to obtain an optimized deep learning model.

[0039] To achieve the above object, the second aspect of the present application provides a wildfire identification method based on unmanned aerial vehicle inspection, which is applied to the wildfire identification system based on unmanned aerial vehicle inspection as described in the first aspect and any feasible implementation manner, and the method comprises the following steps:

[0040] Collecting image data of a power transmission line;

[0041] Identifying wildfire of the image data by using a pre-trained deep learning model, to determine a wildfire identification result of the power transmission line, where the wildfire identification result is used to reflect whether there is wildfire on the power transmission line, and the deep learning model is obtained by pre-training based on historical wildfire image data of the power transmission line;

[0042] Generating a visual prompt based on the wildfire identification result.

[0043] To achieve the above object, the third aspect of the present application provides a wildfire identification device based on unmanned aerial vehicle inspection, which is applied to the wildfire identification system based on unmanned aerial vehicle inspection as described in the first aspect and any feasible implementation manner, and the device comprises:

[0044] A data collection module configured to collect image data of a power transmission line;

[0045] The mountain fire identification module is configured to identify the mountain fire in the image data by using a pre-trained deep learning model, and determine a mountain fire identification result of the power transmission line, wherein the mountain fire identification result is used to reflect whether there is a mountain fire in the power transmission line, and the deep learning model is pre-trained based on historical mountain fire image data of the power transmission line.

[0046] The prompt generation module is configured to generate a visual prompt based on the mountain fire identification result.

[0047] To achieve the above-mentioned purpose, the fourth aspect of the present application provides a computer readable storage medium, which stores a computer program, and the computer program makes the processor execute the steps of the mountain fire identification method based on the unmanned aerial vehicle inspection when executed by the processor.

[0048] To achieve the above-mentioned purpose, the fifth aspect of the present application provides a computer device, which comprises a memory and a processor, and the memory stores a computer program, and the computer program makes the processor execute the steps of the mountain fire identification method based on the unmanned aerial vehicle inspection when executed by the processor.

[0049] By adopting the embodiment of the present application, the following beneficial effects are achieved:

[0050] The present application provides a mountain fire identification system based on unmanned aerial vehicle inspection, which comprises an unmanned aerial vehicle data acquisition system, a deep learning analysis system and a terminal visualization system, and the unmanned aerial vehicle data acquisition system, the deep learning analysis system and the terminal visualization system are communicatively connected; the unmanned aerial vehicle data acquisition system is configured to acquire image data of a power transmission line and send the image data to the deep learning analysis system; the deep learning analysis system is configured to receive the image data, identify the mountain fire in the image data by using a pre-trained deep learning model, determine a mountain fire identification result of the power transmission line, and send the mountain fire identification result to the terminal visualization system, wherein the mountain fire identification result is used to reflect whether there is a mountain fire in the power transmission line, and the deep learning model is pre-trained based on historical mountain fire image data of the power transmission line; and the terminal visualization system is configured to receive the mountain fire identification result and generate a visual prompt based on the mountain fire identification result. Through the above-mentioned mountain fire identification system, an advanced unmanned aerial vehicle inspection system can be provided, which combines a deep learning algorithm to realize real-time monitoring of the power transmission line and its surrounding environment. The advantage of this combined technology is that it can automatically identify potential dangerous situations such as mountain fires and timely issue warnings, thereby significantly improving the safety of the power transmission line and the efficiency of the inspection work, and providing an efficient and intelligent mountain fire automatic identification and alarm system for the power transmission line. This system not only improves the safety management standard of the power transmission line, but also provides strong technical support for the stable operation of the power system. BRIEF DESCRIPTION OF DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced. Obviously, the accompanying drawings in the following description only represent some embodiments of the present application, and for those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0052] In which:

[0053] Figure 1 The structural block of a mountain fire identification system based on unmanned aerial vehicle inspection according to an embodiment of the present application Figure 1 ;

[0054] Figure 2 The structural block of a mountain fire identification system based on unmanned aerial vehicle inspection according to an embodiment of the present application Figure 2 ;

[0055] Figure 3 The structural block of a mountain fire identification system based on unmanned aerial vehicle inspection according to an embodiment of the present application Figure 3 ;

[0056] Figure 4 The structural block of a mountain fire identification system based on unmanned aerial vehicle inspection according to an embodiment of the present application Figure 4 ;

[0057] Figure 5 The flow chart of a mountain fire identification method based on unmanned aerial vehicle inspection according to an embodiment of the present application

[0058] Figure 6 The structural block diagram of a mountain fire identification device based on unmanned aerial vehicle inspection according to an embodiment of the present application

[0059] Figure 1 The structural block diagram of a computer device according to an embodiment of the present application DETAILED DESCRIPTION

[0060] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments only represent some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without any creative effort belong to the scope of protection of the present application.

[0061] The application mainly comprises a background management system and a terminal visualization system, the background management system mainly comprises a UAV data acquisition system 10 and a deep learning analysis system 20, the UAV data acquisition system 10 is equipped with a high-definition camera and a sensor to carry out inspection on a power transmission line, collect image and environmental data, the deep learning analysis system 20 is provided with a deep learning server, a pre-trained deep learning model is used to analyze image data in real time, on this basis, a terminal visualization system 30 is established, is deployed in a line maintenance center of each city, the analysis result of the deep learning analysis module 20 is visually displayed, and auxiliary decision is made.

[0062] Please refer to Figure 1 , Figure 1 The structure of a mountain fire identification system based on UAV inspection according to an embodiment of the application Figure 1 As Figure 2 shown, the mountain fire identification system 00 comprises a UAV data acquisition system 10, a deep learning analysis system 20 and a terminal visualization system 30, wherein the terminal visualization system has a plurality of terminals and is deployed in each city, and the UAV data acquisition system 10, the deep learning analysis system 20 and the terminal visualization system 30 are communicatively connected;

[0063] Further, the unmanned aerial vehicle data acquisition system 10 is configured to acquire image data of the power transmission line and transmit the image data to the deep learning analysis system 20. The image data is image data of the environment in which the power transmission line is located. The image data is acquired in real time, and the existence of a forest fire around the power transmission line can be monitored in real time through the image data. The deep learning analysis system 20 is configured to receive the image data, use a pre-trained deep learning model to identify a forest fire in the image data, determine a forest fire identification result of the power transmission line, and transmit the forest fire identification result to the terminal visualization system 30. The forest fire identification result reflects whether a forest fire exists on the power transmission line. The deep learning model is pre-trained based on historical forest fire image data of the power transmission line. The forest fire identification result can include a first result indicating the existence of a forest fire and a second result indicating the non-existence of a forest fire, as well as corresponding image data, which facilitates intuitive display and prompt of subsequent visualization based on the image data. The historical forest fire image data is image data of the power transmission line acquired when a forest fire occurred at a historical time. The deep learning model is trained through various images when a forest fire occurs, so that the deep learning model can learn the image features of a forest fire. Thus, the occurrence of a forest fire can be analyzed and identified based on real-time acquired image data, and the efficiency and accuracy of forest fire identification are improved. The terminal visualization system 30 is configured to receive the forest fire identification result and generate a visualization prompt based on the forest fire identification result. The forest fire identification result can include the first result or the second result, as well as corresponding image data. The first result and the image data or the second result and the image data are visually displayed, so that relevant personnel can review the forest fire identification result more intuitively and timely.

[0064] By introducing the deep learning algorithm, automatic identification and analysis of forest fire features are realized, and the dependence on manual intervention is reduced. The real-time acquisition mode of the unmanned aerial vehicle has strong real-time performance and can discover a forest fire and issue an alarm in the first time, thereby improving the emergency response speed. The system integrates data analysis and detection functions and can optimize the inspection strategy according to historical data and environmental changes to improve the prevention capability. The system structure design is perfect and can adapt to different environments and weather conditions to ensure the continuity and stability of the inspection work.

[0065] This invention provides a wildfire identification system based on drone inspection. The wildfire identification system includes a drone data acquisition system, a deep learning analysis system, and a terminal visualization system, all interconnected. The drone data acquisition system collects image data of power transmission lines and sends the image data to the deep learning analysis system. The deep learning analysis system receives the image data, uses a pre-trained deep learning model to perform wildfire identification on the image data, determines the wildfire identification result of the power transmission lines, and sends the wildfire identification result to the terminal visualization system. The wildfire identification result reflects whether a wildfire exists on the power transmission lines. The deep learning model is pre-trained based on historical wildfire image data of the power transmission lines. The terminal visualization system receives the wildfire identification result and generates a visual prompt based on the result. This wildfire identification system provides an advanced drone inspection system that combines deep learning algorithms to achieve real-time monitoring of power transmission lines and their surrounding environment. The advantage of this combined technology lies in its ability to automatically identify potential hazards such as wildfires and issue timely warnings, thereby significantly improving the safety of transmission lines and the efficiency of inspection work. It provides transmission lines with a highly efficient and intelligent automatic wildfire identification and alarm system. This system not only raises the safety management standards of transmission lines but also provides strong technical support for the stable operation of the power system.

[0066] Please see Figure 2 , Figure 2 This is a structural framework of a wildfire identification system based on drone inspection, according to an embodiment of the present invention. Figure 2 , Figure 3 The composition of the UAV data acquisition system is specifically demonstrated, including, for example... Figure 3 The UAV data acquisition system 10 shown includes at least: a high-voltage transmission line ledger module 11, a sensor camera control module 12, a data acquisition and storage unit module 13, a data analysis and processing module 14, and a data communication and uploading module 15, wherein the high-voltage transmission line ledger module 11, the sensor camera control module 12, the data acquisition and storage unit module 13, the data analysis and processing module 14, and the data communication and uploading module 15 are connected to each other.

[0067] Further, the high-voltage transmission line account module 11 is used to store the account information of the high-voltage transmission line, and send the account information to the sensor camera control module 12. Specifically, the high-voltage transmission line account module 11 contains the account information of the detailed files of all high-voltage transmission lines, such as the direction of the line, the location of the tower, the specification of the wire, and the related geographical and environmental information. These information is crucial for planning the inspection route, analyzing the risk points, and formulating the maintenance plan. The account module also supports data updating and maintenance, ensuring that all information is up-to-date for effective monitoring and management.

[0068] The sensor camera control module 12 is used to receive the account information, process the data collection planning based on the account information, and obtain the data collection instruction. Based on the data collection instruction, the transmission line is processed for image collection, and the image data is obtained. The image data is sent to the data collection storage unit module 13. Specifically, the sensor camera control module 12 is mainly used to control and coordinate various sensors and cameras on the unmanned aerial vehicle. According to the account information of the high-voltage transmission line account module 11, the high-voltage transmission line to be inspected is automatically collected for photos. This module 12 ensures that during the flight process, data can be collected in the correct way according to the predetermined plan or real-time instruction. It processes image stabilization, focusing, zooming, and sensor switching to optimize data quality and collection efficiency.

[0069] The data collection storage unit module 13 is used to receive and store the image data, and send the image data to the data communication upload module 15. Specifically, the data collection storage unit module 13 is mainly used to receive raw data from sensors and cameras, and store it safely on the memory card or similar storage device of the unmanned aerial vehicle. The data collection storage unit module 13 has high-speed writing capability and large-capacity storage space to adapt to long-time flight tasks and high-resolution video or image storage requirements, and also ensures the integrity of the data to prevent data loss or damage before transmission.

[0070] The data communication upload module 15 is used to receive the image data and upload the image data to the deep learning analysis system 20. Specifically, the data communication upload module 15 mainly uploads the preprocessed transmission line photo information to the subsequent deep learning analysis system 20, providing the system 20 with raw data.

[0071] In a feasible implementation, since the photo information collected by the unmanned aerial vehicle sensors and cameras is affected by weather, environment and the like, the photo data may have large background noise and low signal-to-noise ratio, and therefore, in order to improve the image data quality, the original image data can be subjected to denoising and some preliminary preprocessing operations, and the image data after denoising is sent to the deep learning analysis system 20, that is, the unmanned aerial vehicle data acquisition system 10 can further include a data analysis processing module 14, wherein the input end of the data analysis processing module 14 is in communication connection with the output end of the data acquisition and storage unit module 13, and the output end of the data acquisition and storage unit module 13 is in communication connection with the input end of the data communication uploading module 15;

[0072] Further, the data acquisition and storage unit module 13 is further configured to send the image data to the data analysis processing module 14; the data analysis processing module 14 is configured to receive the image data, perform denoising on the image data to obtain denoised image data, and send the denoised image data to the data communication uploading module 15. It can be understood that, since the photo information collected by the unmanned aerial vehicle sensors and cameras is affected by weather, environment and the like, the photo data may have large background noise and low signal-to-noise ratio, and therefore, the module 14 mainly performs simple preprocessing on the original photo data, mainly including removing background noise, enhancing signal-to-noise ratio and the like, to ensure the effectiveness of the data entering the deep learning analysis system 20 subsequently; and then the data communication uploading module 15 is configured to receive the denoised image data and upload the denoised image data to the deep learning analysis system 20.

[0073] Please refer to Figure 3 , Figure 3 The structure of the forest fire identification system based on unmanned aerial vehicle inspection according to an embodiment of the present application is shown in Figure 4 , Figure 4 The deep learning analysis system 20 is specifically shown in Figure 4 , and at least includes a data preprocessing module 21, a data labeling processing module 22, an image cutting processing module 23, a neural network feature extraction module 24, a model training module 25, a regression prediction module 26 and an output analysis result module 27, wherein the data preprocessing module 21, the data labeling processing module 22, the image cutting processing module 23, the neural network feature extraction module 24, the model training module 25, the regression prediction module 26 and the output analysis result module 27 are in communication connection;

[0074] Further, the data preprocessing module 21 is configured to receive the historical wildfire image data and the image data, preprocess the historical wildfire image data to obtain preprocessed historical wildfire image data, and the preprocessing includes but is not limited to image classification, format standardization, and photo cropping to ensure the consistency of the photo size.

[0075] The data labeling processing module 22 is configured to label the preprocessed historical wildfire image data with training labels to obtain historical wildfire image data labeled with training labels, such as labels of wildfire location, wildfire type, and the like, and to determine the photo category.

[0076] The image cutting processing module 23 is configured to cut the historical wildfire image data labeled with training labels into image blocks according to a preset image cutting rule, and the image cutting rule includes but is not limited to seamless cutting.

[0077] The neural network feature extraction module 24 is configured to extract features of the image blocks by using a neural network model to obtain image features of the image blocks.

[0078] The model training module 25 is configured to train a deep learning model for wildfire recognition by using the image features and the training labels to obtain a trained deep learning model.

[0079] The regression prediction module 26 is configured to perform regression prediction by using the trained deep learning model and the image data to determine a wildfire recognition result of the power transmission line.

[0080] The output analysis result module 27 is configured to send the wildfire recognition result to the terminal visualization system 30.

[0081] The main workflow of the deep learning analysis system 20 is as follows A01 to A08.

[0082] A01, the unmanned aerial vehicle data acquisition system 10 is used to acquire data of the selected high-voltage power transmission line area to establish a massive wildfire detection database, and the elements of the database are power transmission line photos.

[0083] A02, the data preprocessing module 21 is configured to further preprocess the data samples in the database, mainly including image classification, format standardization, and photo cropping to ensure the consistency of the photo size.

[0084] A03, the data labeling processing module 22 is configured to create labels for supervised learning, and manually label the photo data to determine the category of the photo.

[0085] A04. Because the collected and labeled photos are large, the image segmentation processing module 23 performs seamless segmentation on the image sample library dataset to obtain small image blocks, each with a size of 320×320. To achieve seamless segmentation, this invention fills the edges of the image with black pixels during the segmentation process, ensuring that its width and height are both multiples of 320.

[0086] A05. After the segmentation is completed, the obtained small image blocks are fed into the neural network model to complete feature extraction.

[0087] A06. The model training module 25 divides the massive amount of photo data into training and validation sets according to a certain ratio. The training set is used to train the model, and the test set is used to validate the training results. If the accuracy requirement is not met, the number of training set samples is increased to continue training until the accuracy requirement is met, and the trained deep learning model is obtained.

[0088] A07. After the model is trained, the regression prediction module 26 will perform regression prediction on the newly entered photos (such as the image data of power transmission lines collected in real time during actual application) to quickly detect the wildfire area in the image and mark the location of the wildfire in the image with a box to obtain the wildfire recognition result.

[0089] A08, Output Analysis Result Module 27 outputs the wildfire identification results to the terminal visualization system. The wildfire photo information confirmed by the visualization system is added to the training set to continuously train the model and improve the prediction accuracy.

[0090] Please see Figure 4 , Figure 5 This is a structural framework of a wildfire identification system based on drone inspection, according to an embodiment of the present invention. Figure 5 ; Figure 5 The specific structure of the terminal visualization system 30 is illustrated. This system is built upon the UAV data acquisition system 10 and the deep learning analysis system 20, displaying the output results of the deep learning analysis system 20. It is deployed in various municipal-level power transmission line inspection and maintenance centers. The system mainly includes an SMS alarm push module 31, a visualization display module 32, a decision support display module 33, and a feedback loop module 34. These modules are interconnected.

[0091] Further, the short message alarm pushing module 31 is used for receiving the mountain fire identification result, and outputting a short message alarm information to the visual display module 32 according to the mountain fire identification result and the current operation state of the power transmission line. Specifically, the short message alarm pushing module 31 sends an abnormal alarm information to the person in charge of the corresponding city-level power transmission maintenance center according to the detection result, that is, the mountain fire identification result transmitted by the deep learning analysis system, in combination with the power transmission line operation condition, voltage grade, importance and other information, so that the person in charge can master the power transmission line abnormal information in the first time.

[0092] The visual display module 32 is used for receiving the short message alarm information and outputting a visual prompt based on the short message alarm information. Specifically, the person in charge logs into the system quickly after receiving the alarm information, queries the mountain fire photo annotation result corresponding to the alarm information through the visual display module 32, reviews it, and if it is a misjudgment, can click to perform system closed loop to end the task.

[0093] The auxiliary decision display module 33 is used for receiving an artificial auxiliary decision result and sending the auxiliary decision result to the feedback closed loop module 34. The auxiliary decision result is used for indicating the confirmation result of the artificial auxiliary decision on the accuracy of the mountain fire identification result. Specifically, the confirmation result can also include an emergency scheme. If the detection result is correct, the auxiliary decision display module considers the power grid account information, operation mode, and gives auxiliary decision information according to the corresponding power accident event consequence, including processing the mountain fire hidden danger within a specified time or shutting down the line in advance to ensure the safe and stable operation of the power grid.

[0094] The feedback closed loop module 34 is used for receiving the auxiliary decision result, optimizing and training the trained deep learning model as feedback data to obtain an optimized deep learning model. Specifically, after processing the mountain fire information, the operator clicks to complete the closed loop of the mountain fire information, and evaluates the mountain fire detection result and the auxiliary decision information. The evaluation result is fed back to the deep learning analysis module for system optimization.

[0095] The beneficial effects of the present application include at least the following four points:

[0096] 1) The present application collects real-time data by sensors and cameras carried by unmanned aerial vehicles, and uses deep learning algorithms for image and data analysis, which can quickly and accurately identify mountain fires. This method greatly improves the efficiency and accuracy of monitoring, compared with traditional manual patrol or simple sensor-based monitoring systems, it can discover fire in a shorter time, and reduce the delay and misjudgment caused by human factors;

[0097] 2) The application combines a deep learning algorithm, which can realize real-time monitoring of the environment around the power transmission line and automatically identify abnormal conditions such as forest fires. Once signs of forest fires are detected, the system will automatically alert relevant departments or personnel, reducing the response time and providing a valuable time window for decision-making and action in emergency situations, thereby taking preventive measures such as extinguishing or isolating to prevent the spread of fires;

[0098] 3) The automatic monitoring and identification of the application reduces the dependence on manual inspection, saves human resources, and reduces the cost of manual monitoring. At the same time, unmanned aerial vehicle inspection can be carried out in bad weather or complex terrain, reducing the risk of personnel operating in dangerous environments and improving the overall safety of the work;

[0099] 4) The application can automatically adjust the inspection route and frequency according to environmental changes and historical data, optimize the inspection strategy, and ensure the effectiveness and coverage of the monitoring. In addition, data query, statistics, and analysis are carried out through the forest fire management platform, which facilitates scientific data management. All calculations are carried out by software according to unified requirements, with high accuracy and without affecting historical data query statistics and management due to job changes and other factors.

[0100] Please refer to Figures 1 to 4 , Figure 5 The flowchart of a forest fire identification method based on unmanned aerial vehicle inspection in an embodiment of the application is shown in Figure 1 The method can be applied to the forest fire identification system based on unmanned aerial vehicle inspection as described in Figure 1 The method can be applied to the terminal, which can be a desktop terminal or a mobile terminal, and the mobile terminal can be at least one of a mobile phone, a tablet computer, a notebook computer, etc. The server can be implemented by an independent server or a server cluster composed of multiple servers. This embodiment is illustrated by applying to the server. The method comprises the following steps:

[0101] 501, collecting image data of the power transmission line;

[0102] 502, using a pre-trained deep learning model to identify forest fires in the image data, determining the forest fire identification result of the power transmission line, the forest fire identification result reflecting whether there is a forest fire in the power transmission line, and the deep learning model being pre-trained based on historical forest fire image data of the power transmission line;

[0103] 503, generating a visual prompt based on the forest fire identification result.

[0104] It can be understood that Figure 6 The content of each step in the method is the same as that in Figure 6The various systems in the drone-based wildfire identification system shown below function similarly, and will not be elaborated upon here to avoid repetition. For details, please refer to [link / reference needed]. Figure 6 The diagram illustrates the roles of each system in a wildfire identification system based on drone inspection.

[0105] This invention provides a wildfire identification method based on drone inspection. The method includes: collecting image data of power transmission lines; using a pre-trained deep learning model to identify wildfires in the image data, determining the wildfire identification result of the power transmission lines, where the wildfire identification result reflects whether a wildfire exists on the power transmission lines; and generating a visual prompt based on the wildfire identification result. This method provides a wildfire monitoring system combining deep learning algorithms and drone inspection, enabling real-time monitoring of power transmission lines and their surrounding environment. The advantage of this combined technology lies in its ability to automatically identify potential hazards such as wildfires and issue timely warnings, thereby significantly improving the safety of power transmission lines and the efficiency of inspection work. It provides a highly efficient and intelligent automatic wildfire identification and alarm system for power transmission lines. This not only improves the safety management standards of power transmission lines but also provides strong technical support for the stable operation of the power system.

[0106] Please see Figures 1 to 4 , Figure 6 This is a structural block diagram of a wildfire identification device based on drone inspection, as described in an embodiment of the present invention. Figure 5 The device shown is applied to, for example Figure 5 The wildfire identification system based on drone inspection, as described above, includes the following device:

[0107] Data acquisition module 601: Used to acquire image data of power transmission lines;

[0108] Wildfire identification module 602: used to identify wildfires in the image data using a pre-trained deep learning model, and to determine the wildfire identification result of the transmission line. The wildfire identification result is used to reflect whether there is a wildfire on the transmission line. The deep learning model is pre-trained based on the historical wildfire image data of the transmission line.

[0109] Prompt generation module 603: Used to generate visual prompts based on the wildfire identification results.

[0110] Understandable Figure 7 The functions of each module in the system shown are as follows: Figure 7 The steps in the wildfire identification method based on drone inspection shown are similar, and will not be repeated here to avoid repetition. For details, please refer to [link / reference needed]. Figure 7 The diagram shows the steps involved in the wildfire identification method based on drone inspection.

[0111] This invention provides a wildfire identification device based on drone inspection. The device includes: a data acquisition module for acquiring image data of power transmission lines; a wildfire identification module for using a pre-trained deep learning model to identify wildfires in the image data and determine the wildfire identification result of the power transmission line, wherein the wildfire identification result reflects whether a wildfire exists on the power transmission line; and a prompt generation module for generating visual prompts based on the wildfire identification result. This device provides a wildfire monitoring system combining deep learning algorithms and drone inspection, enabling real-time monitoring of power transmission lines and their surrounding environment. The advantage of this combined technology is its ability to automatically identify potential hazards such as wildfires and issue timely warnings, significantly improving the safety of power transmission lines and the efficiency of inspection work. It provides a highly efficient and intelligent automatic wildfire identification and alarm system for power transmission lines. This not only improves the safety management standards of power transmission lines but also provides strong technical support for the stable operation of the power system.

[0112] Figure 5 An internal structural diagram of a computer device in one embodiment is shown. This computer device can specifically be a terminal or a server. Figure 5 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and may also store a computer program, which, when executed by the processor, causes the processor to perform the aforementioned methods. The internal memory may also store a computer program, which, when executed by the processor, causes the processor to perform the aforementioned methods. Those skilled in the art will understand that… ​ The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0113] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform actions such as... ​ The steps of the method shown.

[0114] In one embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, causes the processor to perform the following actions: ​ The steps of the method shown.

[0115] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The program can be stored in a non-volatile computer readable storage medium, and when the program is executed, the processes of the above-mentioned embodiment methods can be included. Any reference to memory, storage, database, or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0116] Any combination of the technical features of the above embodiments can be made. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combination of the technical features does not exist, it should be considered as the scope of the present application.

[0117] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent of the present application. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. A forest fire identification system based on unmanned aerial vehicle inspection, characterized in that, The mountain fire identification system comprises a UAV data acquisition system, a deep learning analysis system and a terminal visualization system, and the UAV data acquisition system, the deep learning analysis system and the terminal visualization system are communicatively connected; The UAV data acquisition system is configured to acquire image data of the power transmission line and send the image data to the deep learning analysis system; The deep learning analysis system is configured to receive the image data, identify the mountain fire of the power transmission line by using a pre-trained deep learning model, determine a mountain fire identification result of the power transmission line, and send the mountain fire identification result to the terminal visualization system, wherein the mountain fire identification result is used to reflect whether there is a mountain fire on the power transmission line, and the deep learning model is pre-trained based on historical mountain fire image data of the power transmission line; The terminal visualization system is configured to receive the mountain fire identification result and generate a visual prompt based on the mountain fire identification result; The deep learning analysis system comprises at least a data preprocessing module, a data labeling processing module, an image cutting processing module, a neural network feature extraction module, a model training module, a regression prediction module and an output analysis result module, and the data preprocessing module, the data labeling processing module, the image cutting processing module, the neural network feature extraction module, the model training module, the regression prediction module and the output analysis result module are communicatively connected; The data preprocessing module is configured to receive the historical mountain fire image data and the image data, preprocess the historical mountain fire image data, and obtain preprocessed historical mountain fire image data; The data labeling processing module is configured to label the preprocessed historical mountain fire image data with training labels, and obtain historical mountain fire image data labeled with training labels; The image cutting processing module is configured to cut the historical mountain fire image data labeled with training labels into image blocks according to a preset image segmentation rule; The neural network feature extraction module is configured to extract features of the image blocks by using a neural network model, and obtain image features of the image blocks; The model training module is configured to train a deep learning model for mountain fire identification by using the image features and the training labels, and obtain a trained deep learning model; The regression prediction module is configured to perform regression prediction by using the trained deep learning model and the image data, and determine the mountain fire identification result of the power transmission line; The output analysis result module is configured to send the mountain fire identification result to the terminal visualization system.

2. The bushfire identification system of claim 1, wherein, The UAV data acquisition system comprises at least a high-voltage power transmission line account module, a sensor camera control module, a data acquisition and storage unit module and a data communication upload module, and the high-voltage power transmission line account module, the sensor camera control module, the data acquisition and storage unit module and the data communication upload module are communicatively connected; The high-voltage power transmission line account module is configured to store account information of the high-voltage power transmission line and send the account information to the sensor camera control module; The sensor camera control module is configured to receive the account information, perform data collection planning processing based on the account information, and obtain a data collection instruction; perform image collection processing on the power transmission line based on the data collection instruction, and obtain the image data; and send the image data to the data collection storage unit module; The data collection storage unit module is configured to receive and store the image data, and send the image data to the data communication upload module; The data communication upload module is configured to receive the image data, and upload the image data to the deep learning analysis system.

3. The bushfire identification system of claim 2, wherein, The unmanned aerial vehicle data collection system further comprises a data analysis processing module, an input end of the data analysis processing module is communicatively connected to an output end of the data collection storage unit module, and the output end of the data collection storage unit module is communicatively connected to an input end of the data communication upload module; The data collection storage unit module is further configured to send the image data to the data analysis processing module; The data analysis processing module is configured to receive the image data, perform denoising processing on the image data, obtain denoised image data, and send the denoised image data to the data communication upload module; The data communication upload module is configured to receive the denoised image data, and upload the denoised image data to the deep learning analysis system.

4. The fire identification system of claim 1, wherein, The terminal visualization system at least comprises a short message alarm pushing module and a visual display module, and the short message alarm pushing module and the visual display module are communicatively connected; The short message alarm pushing module is configured to receive the wildfire identification result, and output short message alarm information to the visual display module according to the wildfire identification result and a current operating state of the power transmission line; The visual display module is configured to receive the short message alarm information, and output a visual prompt based on the short message alarm information.

5. The bushfire identification system of claim 4, wherein, The terminal visualization system further comprises an auxiliary decision display module and a feedback closed loop module, and the auxiliary decision display module and the feedback closed loop module are communicatively connected; The auxiliary decision display module is configured to receive an artificial auxiliary decision result, and send the artificial auxiliary decision result to the feedback closed loop module, the artificial auxiliary decision result being used to indicate a confirmation result of an artificial auxiliary decision on accuracy of the wildfire identification result; The feedback closed loop module is configured to receive the artificial auxiliary decision result, and perform optimization training on the trained deep learning model using the artificial auxiliary decision result as feedback data, to obtain an optimized deep learning model.

6. A forest fire identification method based on unmanned aerial vehicle inspection, characterized in that, The method is applied to the wildfire identification system based on unmanned aerial vehicle inspection according to any one of claims 1 to 5, and the method comprises: collecting image data of a power transmission line; performing wildfire identification on the image data using a pre-trained deep learning model, to determine a wildfire identification result of the power transmission line, the wildfire identification result being used to reflect whether there is a wildfire on the power transmission line, and the deep learning model being pre-trained based on historical wildfire image data of the power transmission line; generate a visual prompt based on the wildfire identification result.

7. A wildfire identification device based on unmanned aerial vehicle inspection, characterized in that, The device is applied to the unmanned aerial vehicle inspection-based wildfire identification system as claimed in any one of claims 1 to 5, and the device comprises: a data collection module, configured to collect image data of a power transmission line; a wildfire identification module, configured to identify a wildfire in the image data by using a pre-trained deep learning model to determine a wildfire identification result of the power transmission line, wherein the wildfire identification result is used to reflect whether there is a wildfire in the power transmission line, and the deep learning model is pre-trained based on historical wildfire image data of the power transmission line; a prompt generation module, configured to generate a visual prompt based on the wildfire identification result.

8. A computer readable storage medium storing a computer program, characterized in that, The computer program is executed by the processor to enable the processor to perform the steps of the method as claimed in claim 6. 9.A computer device, comprising a memory and a processor, and characterized in that, The memory stores a computer program, and the computer program is executed by the processor to enable the processor to perform the steps of the method as claimed in claim 6.

Images