Unmanned aerial vehicle-based discharge port discharge identification method and device, and medium
Through drone identification technology, combined with video frame extraction and data expansion analysis, the accuracy and real-time problems of river outlet identification methods are solved, and real-time monitoring and accurate identification of river outlet status are achieved.
Patent Information
- Application Number
- CN202411335428.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-24
- Publication Date
- 2024-12-27
AI Technical Summary
In the prior art, the river channel outlet identification method cannot accurately and timely determine the river channel outlet status, especially in the case of complex geographical environment and hidden discharge outlets, resulting in insufficient monitoring and real-time monitoring.
Using the drone-based discharge identification method, by obtaining river outlet data, video frame extraction, data expansion analysis, category labeling distinction, drone access identification and focus analysis, combined with regional stage judgment, real-time Monitor the drain water flow status.
Real-time monitoring of the discharge status of the river outlet is realized, the accuracy and real-time detection is improved, and the discharge status can be accurately identified and alarms can be issued in a timely manner.
Smart Images
Figure CN119206551A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of image recognition technology, and in particular to a method, device and medium for identifying discharge outlets based on drones. Background Art
[0002] In the field of river outlet identification and monitoring, traditional supervision methods mainly rely on installing cameras for real-time monitoring and increasing on-site inspections by inspectors. Rivers usually have long extension distances and complex geographical environments, which makes the installation of cameras covering all key areas a difficult task. Traditional camera installation methods not only require a large amount of capital investment, but also involve complex construction and wiring work, especially in river areas with complex terrain and inconvenient transportation, which significantly increases the difficulty and cost of construction.
[0003] In the current river outlet identification and detection methods, it is impossible to accurately and timely determine the status of the river outlet due to the complexity of the river environment and the diversity and varying degrees of concealment of the outlets. Summary of the invention
[0004] The embodiments of the present application provide a method, device and medium for identifying discharge outlets based on drones, which solve the technical problem that the river outlet identification method in the prior art cannot accurately and timely determine the state of the river outlet.
[0005] In a first aspect, an embodiment of the present application provides a method for identifying an outlet discharge based on a drone, characterized in that the method includes: acquiring river outlet data, and based on the river outlet data, extracting frames from a river outlet flow video to obtain river outlet frame data; determining a river outlet data set based on the river outlet frame data through frame data expansion analysis; classifying and distinguishing the river outlet data set to obtain a river outlet training data set; based on the river outlet training data set, determining real-time identification data of the river outlet through drone access identification; performing drone focusing analysis on the real-time identification data of the river outlet to determine the outlet water flow data; and determining the outlet water flow state through regional stage judgment based on the outlet water flow data.
[0006] In one implementation of the present application, based on the river outlet data, the river outlet frame data is obtained by extracting frames from the river outlet flow video, specifically including: performing river parameter analysis on the river outlet data to obtain frame extraction influencing parameters; based on the frame extraction influencing parameters, extracting a single frame number of a preset time period of the time video stream data to obtain river channel related frame data; performing outlet detection on the river channel related frame data to determine the river outlet frame data.
[0007] In one implementation of the present application, a river outlet data set is determined based on the river outlet frame data through frame data expansion analysis, specifically including: supplementing the river outlet frame data with a public data set to obtain a first river outlet data set; when the first river outlet data set cannot meet the preset data set satisfaction, using a generative adversarial network method to expand the first river outlet data set to obtain a second river outlet data set; and determining the river outlet data set based on the first river outlet data set and / or the second river outlet data set.
[0008] In one implementation of the present application, a river outlet data set is classified and labeled to obtain a river outlet training data set, specifically including: based on the river outlet data, river outlet annotation data is determined through river outlet annotation; wherein the river outlet annotation data includes: outlet annotation, water flow annotation; the river outlet annotation data is divided into data sets to obtain a river outlet training data set; wherein the river outlet training data set includes: training set, test set, and validation set.
[0009] In one implementation of the present application, based on a river outlet training data set, the real-time identification data of the river outlet is determined through drone access identification, which specifically includes: inputting the river outlet training data set into a preset river outlet identification model to be trained, until the river outlet identification model to be trained converges, and determining the river outlet identification model; setting the river outlet identification model in a preset edge AI gateway, and connecting it to the identification drone through real-time streaming; photographing the corresponding river channel by the identification drone to obtain real-time data of the river channel to be analyzed; performing an outlet suspicion analysis on the real-time data of the river channel to be analyzed, and when the result of the outlet suspicion analysis is that there is no suspected outlet, the video frame corresponding to the no suspected outlet is set to no outlet; when the result of the outlet suspicion analysis is that there is a suspected outlet, continuous frame identification is performed on the frame image after the video frame corresponding to the suspected outlet, and when the video frame corresponding to the suspected outlet obtained by the continuous frame identification is greater than a preset threshold, the real-time identification data of the river outlet is determined; wherein the real-time identification data of the river outlet includes: real-time identification data of the river channel, and real-time identification data of the outlet water flow.
[0010] In one implementation of the present application, a drone focusing analysis is performed on the real-time identification data of the river outlet to determine the outlet water flow data, specifically including: determining the drone hovering coordinates based on the real-time identification data of the river outlet; determining the drone focal length change through drone focusing variable analysis based on the drone hovering coordinates; determining the outlet water flow data through drone hovering identification based on the drone focal length change.
[0011] In one implementation of the present application, the outlet water flow state is determined according to the outlet water flow data through regional stage judgment, specifically including: determining the judgment time period based on a preset time stage threshold; performing outlet existence identification on the outlet water flow data according to the judgment time period, and when the result of the outlet existence identification is that only the outlet exists, the outlet water flow state is set to no discharge; when the result of the outlet existence identification is that there is an outlet and water flow, judging the array logic of the outlet and the water flow, and when the array logic is true, the outlet water flow state is set to the outlet is discharging; when the array logic is false, the outlet water flow state is set to the outlet only exists.
[0012] In one implementation of the present application, after determining the outlet water flow state through regional stage judgment based on the outlet water flow state, the method also includes: determining a communication instruction based on the outlet water flow state, and pushing the communication instruction to confirm the communication point to obtain alarm communication data; obtaining the location information of the outlet water flow state, and sending the alarm picture and video clip corresponding to the outlet water flow state to a preset alarm system.
[0013] In the second aspect, the embodiment of the present application also provides an outlet discharge identification device based on a drone, characterized in that the device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can: obtain river outlet data, and based on the river outlet data, extract frames from the river outlet flow video to obtain river outlet frame data; determine the river outlet data set through frame data expansion analysis based on the river outlet frame data; perform category labeling and differentiation on the river outlet data set to obtain a river outlet training data set; determine the river outlet real-time identification data through drone access identification based on the river outlet training data set; perform drone focusing analysis on the river outlet real-time identification data to determine the outlet water flow data; determine the outlet water flow state through regional stage judgment based on the outlet water flow data.
[0014] In a third aspect, an embodiment of the present application also provides a non-volatile computer storage medium for drone-based outlet discharge identification, storing computer executable instructions, characterized in that the computer executable instructions are set to: obtain river outlet data, and based on the river outlet data, extract frames through the river outlet flow video to obtain river outlet frame data; determine the river outlet data set through frame data expansion analysis based on the river outlet frame data; perform category labeling and differentiation on the river outlet data set to obtain a river outlet training data set; based on the river outlet training data set, determine the river outlet real-time identification data through drone access identification; perform drone focusing analysis on the river outlet real-time identification data to determine the outlet water flow data; determine the outlet water flow state through regional stage judgment based on the outlet water flow data.
[0015] The embodiments of the present application provide a method, device and medium for identifying discharge outlets based on drones. Through the algorithm of the discharge outlet water flow and the hovering identification strategy of the drone, the technical problem that the river outlet identification method in the prior art cannot accurately and timely determine the river outlet status, realizes real-time monitoring of the discharge status of the river outlet, and improves the accuracy and real-time performance of the river outlet status detection. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0017] Figure 1 A method flow chart provided for an embodiment of the present application;
[0018] Figure 2 A schematic diagram of the internal structure of a device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0019] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.
[0020] The embodiments of the present application provide a method, device and medium for identifying discharge outlets based on drones. Through the algorithm of the discharge outlet water flow and the hovering identification strategy of the drone, the technical problem that the river outlet identification method in the prior art cannot accurately and timely determine the river outlet status, realizes real-time monitoring of the discharge status of the river outlet, and improves the accuracy and real-time performance of the river outlet status detection.
[0021] The technical solution proposed in the embodiments of the present application is described in detail below with reference to the accompanying drawings.
[0022] Figure 1 A flow chart of a method for identifying an outlet discharge based on a drone is provided in an embodiment of the present application. Figure 1 As shown, an outlet emission identification method based on a drone provided in an embodiment of the present application specifically includes the following steps:
[0023] Step 101, obtaining river outlet data, and based on the river outlet data, extracting frames through the river outlet flow video to obtain river outlet frame data;
[0024] Specifically, the method includes: performing river channel parameter analysis on the river channel outlet data to obtain frame extraction influencing parameters; based on the frame extraction influencing parameters, extracting a single frame number in a time period of a preset time video stream data to obtain river channel related frame data; and performing outlet detection on the river channel related frame data to determine the river channel outlet frame data.
[0025] The present application obtains river outlet data, and based on the river outlet data, extracts frames from the river outlet flow video to obtain river outlet frame data, thereby achieving frame data acquisition for the river outlet video stream and improving the efficiency of river outlet detection.
[0026] In the examples of the present application, detailed explanation is given by the following Example 1.
[0027] Example 1: By planning the route of the drone along the centerline of the river, the flight altitude is controlled within 10-20 meters from the water surface, and the camera angle is inclined at 45° to the heading, the river outlet data is collected.
[0028] According to the collected river outlet data and / or historical video data, the frame extraction interval is analyzed by parameters, and the frame extraction interval is expressed by the following formula:
[0029]
[0030] Wherein, A represents the frame extraction interval;
[0031] a is the width of the camera’s field of view;
[0032] v is the heading speed of the UAV;
[0033] F is the video stream frame rate.
[0034] The river outlet data is processed by frame extraction using the obtained frame extraction interval A to obtain river channel related frame data.
[0035] Finally, the images with outlet information are screened out to determine the river outlet frame data.
[0036] Step 102: Determine a river outlet data set based on the river outlet frame data by frame data expansion analysis.
[0037] Specifically, the method includes: supplementing the river outlet frame data with a public data set to obtain a first river outlet data set; when the first river outlet data set cannot meet the preset data set satisfaction, using a generative adversarial network method to expand the first river outlet data set to obtain a second river outlet data set; and determining the river outlet data set based on the first river outlet data set and / or the second river outlet data set.
[0038] This application determines the river outlet data set based on the river outlet frame data through frame data expansion analysis, thereby achieving data set expansion when the data set is insufficient.
[0039] In the examples of the present application, detailed explanation is given by the following Example 2.
[0040] Example 2: The river outlet frame data is supplemented with public data sets to obtain the first river outlet data set.
[0041] The first river outlet dataset has a relatively small amount of data and cannot be supplemented by public image datasets. The data is expanded through the Generative Adversarial Network (GAN) to obtain the second river outlet dataset.
[0042] Generative adversarial networks, or GANs, are divided into generators and discriminators. Both are neural network models. The task of the generator network is to generate distributions based on a set of random noise data to form new data; the task of the discriminator network is to evaluate whether the newly generated data is real, that is, whether it comes from a real training set. The GAN network can expand data in large quantities, and the image clarity can also be enhanced. GAN is highly operational and can learn various data features without supervision.
[0043] Step 103: classify and label the river outlet data set to obtain a river outlet training data set.
[0044] Specifically including: based on the river outlet data, through the river outlet marking, determining the river outlet marking data; wherein the river outlet marking data includes: outlet marking, water flow marking; dividing the river outlet marking data into data sets to obtain a river outlet training data set; wherein the river outlet training data set includes: training set, test set, and verification set.
[0045] The present application classifies and labels the river outlet data set to obtain a river outlet training data set, thereby achieving class labeling of the river outlet training data set and improving the accuracy and training efficiency of the river outlet training data set training.
[0046] In the examples of the present application, this is explained in detail through the following Example 3.
[0047] Example 3: Classify and label the established outlet data set, set two labels including gate (outlet) and water (water flow), and complete the classification of training set, test set, and validation set.
[0048] Through label annotation, the river outlet training data set can meet the recognition accuracy of the river outlet recognition model for different pixels and scenes.
[0049] Step 104: Based on the river outlet training data set, the real-time identification data of the river outlet is determined by accessing and identifying the drone.
[0050] Specifically, it includes: inputting the river outlet training data set into the preset river outlet recognition model to be trained, until the river outlet recognition model to be trained converges, and determining the river outlet recognition model; setting the river outlet recognition model in a preset edge AI gateway, and connecting it to the recognition drone through real-time streaming; photographing the corresponding river through the recognition drone to obtain the real-time data of the river to be analyzed; performing an outlet suspicion degree analysis on the real-time data of the river to be analyzed, and when the outlet suspicion degree analysis result is that there is no suspected outlet, setting the video frame corresponding to the no suspected outlet as no outlet; when the outlet suspicion degree analysis result is that there is a suspected outlet, performing continuous frame recognition on the frame image after the video frame corresponding to the suspected outlet, and when the video frame corresponding to the suspected outlet obtained by the continuous frame recognition is greater than a preset threshold, determining the river outlet real-time recognition data; wherein, the river outlet real-time recognition data includes: river real-time recognition data, outlet water flow real-time recognition data.
[0051] This application is based on the river outlet training data set. Through drone access and identification, the real-time identification data of the river outlet is determined, which realizes the real-time data acquisition and analysis of the river outlet, and improves the accuracy and efficiency of river outlet identification.
[0052] In the examples of the present application, a detailed explanation is given by the following Example 4.
[0053] Example 4: Use Yolov8 to train and debug the established data set to obtain a trained model.
[0054] Since the discharge event at the outlet lasts for a long time and the detection frequency requirement is relatively low, and considering that the deployment and operation environment is NVIDIA Jetson smart device edge deployment, after comprehensive comparison, the model used in this application is Yolov8M.
[0055] The trained model is deployed on the edge AI gateway, and the real-time stream of the drone is connected to the trained outlet model, which will output the detected outlet and water flow information in real time.
[0056] Specifically, the outlet model is used to identify whether there are suspected outlets in the river channel images taken by the drone; if a suspected outlet is detected in a certain frame of image, it is determined whether several subsequent frames of the image also detect suspected outlet categories. When the number of outlet categories detected in several consecutive frames of images is greater than the preset number, it is determined that there is an outlet in the current scene.
[0057] It should be noted that for the detection frequency characteristics of the data analysis of pollution drones on river outlets, the best model used in the experiment is Yolov8M.
[0058] Step 105: Perform drone focusing analysis on the real-time identification data of the river outlet to determine the outlet water flow data.
[0059] Specifically include: determining the drone hovering coordinates based on the real-time identification data of the river outlet; determining the drone focal length change based on the drone hovering coordinates through drone focusing variable analysis; determining the outlet water flow data through drone hovering identification based on the drone focal length change.
[0060] This application performs drone focusing analysis on real-time river outlet identification data to determine outlet water flow data, thereby achieving real-time and accurate identification of river outlets and improving the accuracy and efficiency of river outlet identification.
[0061] In the examples of the present application, this is explained in detail through the following Example 5.
[0062] Example 5: When the drone is performing dynamic flight, it will acquire and detect the river outlet in real time, which will cause a certain blur due to the flight speed, so that the pixels in the acquired video and image will be different from the actual pixel arrangement. Therefore, it is necessary to intersperse hovering and zooming during the flight of the drone to further improve the recognition accuracy of the river outlet.
[0063] The adjustment value for the focal length of the drone camera is expressed by the following formula:
[0064]
[0065] Where Δf is the required adjustment value of the drone’s camera focal length;
[0066] D is the number of pixels corresponding to the width of the row opening detected in the previous frame image;
[0067] f0 represents the initial focal length of the drone camera;
[0068] m represents the number of pixels in each row of the image taken by the drone;
[0069] L represents the actual average width of the river channel.
[0070] Step 106: Determine the outlet water flow state through regional stage determination based on the outlet water flow data.
[0071] Specifically, it includes: determining a judgment time period based on a preset time stage threshold; identifying the existence of an outlet for the outlet water flow data according to the judgment time period, and when the result of the outlet existence identification is that only an outlet exists, the outlet water flow state is set to no discharge; when the result of the outlet existence identification is that there is an outlet and water flow, judging the array logic of the outlet and the water flow, and when the array logic is true, the outlet water flow state is set to that the outlet is discharging; when the array logic is false, the outlet water flow state is set to that only an outlet exists.
[0072] After determining the outlet water flow state through regional stage judgment according to the outlet water flow state, the method also includes: determining a communication instruction based on the outlet water flow state, and pushing the communication instruction to confirm the communication point to obtain alarm communication data; obtaining the location information of the outlet water flow state, and sending the alarm picture and video clip corresponding to the outlet water flow state to a preset alarm system.
[0073] This application determines the outlet water flow status based on the outlet water flow data through regional stage judgment, realizes the accurate judgment of the real-time status of the river outlet in the environment of drone and recognition algorithm, and improves the accuracy of river outlet status identification.
[0074] In the examples of the present application, this is explained in detail through the following Example 6.
[0075] Example 6: Determine whether the outlet is discharging: Based on the output outlet and water flow information, input the AI outlet status determination algorithm to determine and output the current outlet status.
[0076] The specific judgment algorithm is:
[0077] If the only result obtained in the judgment area is gate within a certain period of time, it is judged as no discharge;
[0078] If the region recognizes a gate (outlet) and has a water (water flow) category, and the logical operation of the two category arrays is true, then it is determined that the outlet is discharging;
[0079] If the region recognizes a gate (outlet) and has a water (water flow) category, and the logical operation of the two category arrays is false, it is determined that only the outlet is detected.
[0080] When receiving a notification that the outlet is discharging, a text message notification is sent to the person responsible for the current point, and the current real-time location information is obtained through the GPS carried by the drone. At the same time, the warning pictures and video clips captured at the current point are pushed to the alarm system carried by the edge AI gateway.
[0081] The above is an embodiment of the method proposed in this application. Based on the same inventive concept, the embodiment of this application also provides a discharge outlet identification device based on a drone, and its structure is as follows: Figure 2 shown.
[0082] Figure 2 A schematic diagram of the internal structure of a discharge outlet identification device based on a drone provided in an embodiment of the present application. Figure 2 As shown, the device includes:
[0083] at least one processor 201;
[0084] and, a memory 202 communicatively connected to the at least one processor;
[0085] The memory 202 stores instructions that can be executed by at least one processor, and the instructions are executed by at least one processor 201 to enable at least one processor 201 to:
[0086] The river outlet data is obtained, and based on the river outlet data, frame data of the river outlet is obtained by extracting frames from the river outlet flow video; based on the river outlet frame data, a river outlet data set is determined through frame data expansion analysis; the river outlet data set is categorized and distinguished to obtain a river outlet training data set; based on the river outlet training data set, the real-time identification data of the river outlet is determined through drone access identification; the real-time identification data of the river outlet is analyzed by drone focusing to determine the outlet water flow data; based on the outlet water flow data, the outlet water flow state is determined through regional stage judgment.
[0087] Some embodiments of the present application provide corresponding Figure 1 A non-volatile computer storage medium for unmanned aerial vehicle-based discharge port emission identification stores computer executable instructions, wherein the computer executable instructions are set to:
[0088] The river outlet data is obtained, and based on the river outlet data, frame data of the river outlet is obtained by extracting frames from the river outlet flow video; based on the river outlet frame data, a river outlet data set is determined through frame data expansion analysis; the river outlet data set is categorized and distinguished to obtain a river outlet training data set; based on the river outlet training data set, the real-time identification data of the river outlet is determined through drone access identification; the real-time identification data of the river outlet is analyzed by drone focusing to determine the outlet water flow data; based on the outlet water flow data, the outlet water flow state is determined through regional stage judgment.
[0089] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the IoT device and medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.
[0090] The system and medium provided in the embodiments of the present application correspond one-to-one to the method. Therefore, the system and medium also have similar beneficial technical effects to the corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the system and medium will not be repeated here.
[0091] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
[0092] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0093] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0094] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0095] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0096] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.
[0097] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0098] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0099] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.
Claims
1. A method for identifying discharge outlets based on drones, characterized in that: The method comprises: Acquire river outlet data, and based on the river outlet data, extract frames from the river outlet flow video to obtain river outlet frame data; Determine a river outlet data set through frame data expansion analysis based on the river outlet frame data; Classifying and distinguishing the river outlet data set to obtain a river outlet training data set; Based on the river outlet training data set, the real-time identification data of the river outlet is determined by accessing and identifying the drone; Performing drone focusing analysis on the real-time identification data of the river outlet to determine the outlet water flow data; The outlet water flow state is determined based on the outlet water flow data through regional stage judgment.
2. The method for identifying discharge outlets based on drones according to claim 1, characterized in that: Based on the river outlet data, the river outlet frame data is obtained by extracting frames from the river outlet flow video, specifically including: Performing a river channel parameter analysis on the river channel outlet data to obtain frame extraction influencing parameters; Based on the frame extraction influencing parameters, single frames are extracted from the preset time video stream data in a time period to obtain river channel related frame data; The outlet detection is performed on the river channel related frame data to determine the river channel outlet frame data.
3. The method for identifying discharge outlets based on drones according to claim 1, characterized in that: According to the river outlet frame data, a river outlet data set is determined through frame data expansion analysis, specifically including: Supplementing the river outlet frame data with a public data set to obtain a first river outlet data set; When the first river outlet dataset fails to satisfy the preset dataset satisfaction, the first river outlet dataset is expanded by a generative adversarial network method to obtain a second river outlet dataset; The river channel outlet dataset is determined according to the first river channel outlet dataset and / or the second river channel outlet dataset.
4. The method for identifying discharge outlets based on drones according to claim 1, characterized in that: The river outlet data set is classified and marked to obtain a river outlet training data set, which specifically includes: Based on the river outlet data, river outlet marking data is determined through river outlet marking; wherein the river outlet marking data includes: outlet marking and water flow marking; The river outlet annotation data is divided into data sets to obtain the river outlet training data set; wherein the river outlet training data set includes: a training set, a test set, and a validation set.
5. The method for identifying discharge outlets based on drones according to claim 1, characterized in that: Based on the river outlet training data set, the real-time identification data of the river outlet is determined by accessing and identifying the drone, specifically including: Inputting the river outlet training data set into a preset river outlet recognition model to be trained, until the river outlet recognition model to be trained converges, and determining the river outlet recognition model; The river outlet recognition model is set in a preset edge AI gateway and connected to the recognition drone through real-time streaming; The corresponding river channel is photographed by the identification drone to obtain real-time data of the river channel to be analyzed; Performing an outlet suspicion degree analysis on the real-time data of the river channel to be analyzed, and when the outlet suspicion degree analysis result is that there is no suspected outlet, setting the video frame corresponding to the no suspected outlet as no outlet; When the outlet suspicion analysis result is that there is a suspected outlet, continuous frame recognition is performed on the frame image after the video frame corresponding to the suspected outlet, and when the video frame corresponding to the suspected outlet obtained by the continuous frame recognition is greater than a preset threshold, the real-time recognition data of the river outlet is determined; wherein the real-time recognition data of the river outlet includes: real-time identification data of the river channel and real-time identification data of the outlet water flow.
6. The method for identifying discharge outlets based on drones according to claim 1, characterized in that: The real-time identification data of the river outlet is subjected to drone focusing analysis to determine the outlet water flow data, specifically including: Determining the hovering coordinates of the drone based on the real-time identification data of the river outlet; According to the hovering coordinates of the drone, determining the focal length change of the drone through the drone focusing variable analysis; Based on the change in focal length of the drone, the outlet water flow data is determined through drone hovering recognition.
7. The method for identifying discharge outlets based on drones according to claim 1, characterized in that: According to the outlet water flow data, the outlet water flow state is determined through regional stage determination, specifically including: Determine a determination time period based on a preset time stage threshold; According to the judgment time period, the outlet water flow data is subjected to outlet existence identification, and when the outlet existence identification result is that only the outlet exists, the outlet water flow state is set to no discharge; When the result of the outlet existence identification is that there is an outlet and a water flow, the array logic of the outlet and the water flow is judged, and when the array logic is true, the outlet water flow state is set to the outlet being discharged; When the array logic is false, the outlet water flow state is set to only have an outlet.
8. The method for identifying discharge outlets based on drones according to claim 1, characterized in that: After determining the outlet water flow state through regional stage determination according to the outlet water flow state, the method further includes: Based on the outlet water flow state, determine the communication instruction, and push the communication instruction to confirm the push point to obtain the alarm communication data; The location information of the outlet water flow state is obtained, and the warning picture and video clip corresponding to the outlet water flow state are sent to a preset warning system.
9. A discharge outlet identification device based on a drone, characterized in that: The device comprises: at least one processor; and, a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: Acquire river outlet data, and based on the river outlet data, extract frames from the river outlet flow video to obtain river outlet frame data; Determine a river outlet data set through frame data expansion analysis based on the river outlet frame data; Classifying and distinguishing the river outlet data set to obtain a river outlet training data set; Based on the river outlet training data set, the real-time identification data of the river outlet is determined by accessing and identifying the drone; Performing drone focusing analysis on the real-time identification data of the river outlet to determine the outlet water flow data; The outlet water flow state is determined based on the outlet water flow data through regional stage judgment.
10. A non-volatile computer storage medium for drone-based discharge identification, storing computer executable instructions, characterized in that: The computer executable instructions are configured to: Acquire river outlet data, and based on the river outlet data, extract frames from the river outlet flow video to obtain river outlet frame data; Determine a river outlet data set through frame data expansion analysis based on the river outlet frame data; Classifying and distinguishing the river outlet data set to obtain a river outlet training data set; Based on the river outlet training data set, the real-time identification data of the river outlet is determined by accessing and identifying the drone; Performing drone focusing analysis on the real-time identification data of the river outlet to determine the outlet water flow data; The outlet water flow state is determined based on the outlet water flow data through regional stage judgment.