Beach garbage identification and classification method and coastal pollution early warning system
Through the beach garbage identification method combined with three-dimensional lidar and drone image data, the efficiency of beach garbage identification and coastal pollution monitoring is solved, efficient identification of beach garbage and real-time monitoring of coastal pollution levels are realized, and the coastal environment is improved Clean management efficiency.
Patent Information
- Application Number
- CN202411248537.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-06
- Publication Date
- 2024-11-29
AI Technical Summary
The existing technology lacks efficient and accurate beach garbage identification and classification methods and coastal pollution automatic warning systems, resulting in the coastal environmental cleaning strategy being not fast and effective enough, making it difficult to monitor the type and quantity of beach garbage in real time and determine the pollution level.
Three-dimensional lidar scanning is used to obtain terrain and garbage point cloud data, combine Smart3D software to reconstruct the terrain model, obtain real-time image data input through drones for garbage identification, and calculate coastal pollution levels based on meteorological data to generate early warnings hint.
It realizes efficient identification and classification of beach garbage and real-time monitoring of coastal pollution, reduces the cost error of manual investigation, and improves the efficiency of coastal environmental pollution monitoring and cleaning.
Smart Images

Figure CN119049039A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of marine environment monitoring, and in particular to a beach garbage identification and classification method and a coastal pollution early warning system. Background Art
[0002] Marine litter refers to items that have been discarded, misplaced, or carried to the ocean and coastline by wind and rivers. They are mainly composed of plastics, metals, glass, rubber, fabrics, and paper. Land-based litter accounts for 80% of marine litter, with tourism activities, illegal sewage discharges, and poorly managed landfills being the main contributors. Marine litter is found in all marine ecological environments, from urban beaches to deep sea trenches, from coasts, beaches, cliffs, dunes, and ports to the surface of the water where garbage floats, to the high seas and mid-oceans. Plastic is the most common type of litter in coastal areas, and this polymer has become a major threat to the marine environment due to its chemical composition and biodegradation characteristics. Plastic litter can take hundreds of years to break down and disappear. Plastic is widely used because it is cheap, easy to produce, and lasts a long time. It is considered an indispensable part of modern life. As the world's demand for plastics continues to grow, with recent estimates that plastic production has increased by 43% in the past decade, plastic waste management remains a global challenge and is now considered a major threat to marine biota.
[0003] At present, there is an urgent need for an efficient and accurate beach garbage identification and classification method and an automatic coastal pollution early warning system. By real-time monitoring the type and amount of beach garbage and determining the pollution level of the coast, early warning alarms are generated to remind local governments and beach management departments, so as to formulate faster and more effective garbage cleaning strategies to ensure the cleanliness of the coastal environment. Therefore, a beach garbage identification and classification method and a coastal pollution early warning system are proposed. Summary of the invention
[0004] The purpose of the present invention is to provide a beach garbage identification and classification method and a coastal pollution early warning system to solve the problems raised in the above background technology.
[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:
[0006] In a first aspect, the present invention provides a method for identifying and classifying beach garbage, comprising the following steps:
[0007] S1: Scan the monitoring area through 3D laser radar to obtain 3D terrain coordinates and garbage point cloud data;
[0008] S2: Reconstruct the terrain model of the monitoring area through modeling software, calculate the area of the monitoring area, and determine which beach type the area belongs to, and transmit the results to the comprehensive judgment module;
[0009] S3: Calculate the amount and distribution of garbage in the monitoring area based on the garbage point cloud data, and preliminarily determine whether there is garbage. If there is garbage, the amount of garbage is transmitted to the comprehensive judgment module. If there is no garbage, only the terrain judgment result is transmitted to the comprehensive judgment module.
[0010] S4: The real-time image data of the monitoring area is obtained through the drone, and input into the trained neural network model to identify whether there are garbage targets in the image, and count the amount of garbage, and transmit the identification and classification results and garbage statistics results to the comprehensive judgment module;
[0011] S5: In the comprehensive determination module, the clean coast score is calculated according to the input data corresponding to each indicator, and the coastal pollution level is determined at the same time, and the determination result is input into the early warning module;
[0012] S6: In the early warning module, an early warning is generated according to the coastal pollution level and published to the network terminal.
[0013] A further improvement of the technical solution of the present invention is that the modeling software in S2 is Smart3D software.
[0014] A further improvement of the technical solution of the present invention is that the point cloud data acquisition software in S3 is a laser radar scanner.
[0015] A further improvement of the technical solution of the present invention is that the process of establishing the trained neural network model in S4 includes:
[0016] Step 1: Use digital cameras and drones to take pictures of beach garbage targets to build a data set;
[0017] Step 2: Divide the data set into training set and test set in a ratio of 8:2;
[0018] Step 3: Input the training set into the built neural network model, use back propagation to adjust parameters such as the number of iterations and learning rate, and continue training until the loss function value reaches the minimum, obtain the optimal parameters at this time, and save the optimized model;
[0019] Step 4: Input the test set into the trained optimization model for testing, and then use the average detection accuracy to evaluate the detection effect of the model.
[0020] A further improvement of the technical solution of the present invention is that the coastal pollution level is determined in S5 using a coastal pollution level discrimination table.
[0021] A further improvement of the technical solution of the present invention is that the calculation of the coastal pollution level discrimination table is to assign weights to each indicator in the coastal pollution level discrimination table and calculate the final total score TP.
[0022] In a second aspect, the present invention also provides a coastal pollution early warning system, including a terrain data acquisition module, a preliminary judgment module, a meteorological data acquisition module, an image data acquisition module, a network identification module, and a comprehensive judgment module;
[0023] Terrain data acquisition module: used to obtain the three-dimensional terrain coordinate data of the monitoring area, reconstruct the three-dimensional terrain through Smart3D software, and determine the type of beach in the monitoring area, and then transmit the results to the comprehensive judgment module and early warning module;
[0024] Preliminary judgment module: Preliminary judgment on whether there is garbage based on the beach garbage point cloud data in the terrain acquisition module, and transmit the judgment result to the comprehensive judgment module;
[0025] Meteorological data acquisition module: used to obtain the wind direction and wind speed in the monitoring area and transmit the data to the comprehensive judgment module;
[0026] Image data acquisition module: used to obtain digital images of coastal garbage in the monitoring area and transmit the images to the network recognition module;
[0027] Network recognition module: inputs the digital image passed by the image acquisition module into the pre-trained neural network model to detect whether there is beach garbage in the image, and passes the recognition result to the comprehensive judgment module;
[0028] Comprehensive judgment module: sort out the input data, calculate the coastal pollution level of the monitoring area through the above-mentioned coastal pollution level judgment table, and output the judgment result to the network terminal in real time;
[0029] Early warning module: According to the pollution level judgment result output by the comprehensive judgment module, a real-time early warning is issued. If the level is very clean, the network terminal displays a dark green prompt; if the level is clean, the network terminal displays a light green prompt; if the level is medium clean, the network terminal displays a yellow warning; if the level is dirty, the network terminal displays an orange warning; if the level is very dirty, the network terminal displays a red warning and issues an alarm.
[0030] In a third aspect, the present invention provides a terminal device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements any of the above-mentioned beach garbage identification and classification methods and coastal pollution warning systems when executing the computer program.
[0031] In a fourth aspect, a computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any of the beach garbage identification and classification methods and coastal pollution early warning systems described above.
[0032] Due to the adoption of the above technical solution, the present invention has the following technical advances compared with the prior art:
[0033] The present invention provides a beach garbage identification and classification method and a coastal pollution early warning system. The present invention collects data from monitoring beaches and uses a pre-trained neural network structure to detect garbage in real-time images taken by drones on the beach. The coastal pollution level is comprehensively judged based on factors such as beach type and characteristics, wind speed, wind direction, and the presence or absence of garbage cleaners, providing a new method for beach garbage identification and classification, coastal cleanliness evaluation, and pollution early warning, which can greatly reduce the cost error of manual investigation and improve the monitoring and cleaning efficiency of coastal environmental pollution. BRIEF DESCRIPTION OF THE DRAWINGS
[0034] Figure 1 A flowchart for a beach garbage identification and classification method and a coastal pollution identification system;
[0035] Figure 2 It is a flow chart of the connection between modules in a beach garbage identification and classification method and a coastal pollution identification system;
[0036] Figure 3 It is a flow chart of a beach garbage identification and classification method and a neural network model training process in a coastal pollution identification system;
[0037] Figure 4 Construct a flow chart for a beach trash identification and classification method and a neural network model for a coastal pollution discrimination system;
[0038] Figure 5 This is a schematic diagram of a beach garbage identification and classification method and a pollution level determination system for a coastal pollution identification system. DETAILED DESCRIPTION
[0039] The present invention is further described in detail below in conjunction with embodiments:
[0040] First, as Figure 1-5 As shown, the present invention provides a method for identifying and classifying beach garbage, comprising the following steps:
[0041] S1: Scan the monitoring area through 3D laser radar to obtain 3D terrain coordinates and garbage point cloud data;
[0042] S2: Reconstruct the terrain model of the monitoring area through Smart3D software, calculate the area of the monitoring area, and determine which beach type the area belongs to, and transmit the results to the comprehensive judgment module;
[0043] S3: Based on the amount and distribution of garbage in the monitoring area calculated from the garbage point cloud data, a preliminary judgment is made as to whether there is garbage. If there is garbage, the amount of garbage is transmitted to the comprehensive judgment module. If there is no garbage, only the beach type judgment result is transmitted to the comprehensive judgment module.
[0044] S4: Obtain the real-time wind speed and wind direction data of the monitoring area through the meteorological station and transmit it to the comprehensive judgment module;
[0045] S5: The real-time image data of the monitoring area is obtained through the camera, and input into the trained neural network model to identify whether there is garbage in the image, and transmit the identification and classification results to the comprehensive judgment module. The establishment process of the trained neural network model includes:
[0046] Step 1: Use digital cameras and drones to take pictures of garbage targets in advance to build a data set;
[0047] Step 2: Divide the data set into training set and test set in a ratio of 8:2;
[0048] Step 3: Input the training set into the built neural network model, use back propagation to adjust parameters such as the number of iterations and learning rate, and continue training until the loss function value reaches the minimum, obtain the optimal parameters at this time, and save the optimized model;
[0049] Step 4: Input the test set into the trained optimization model for testing, and then use the average detection accuracy to evaluate the detection effect of the model;
[0050] S6: In the comprehensive judgment module, according to the input data corresponding to each indicator, the comprehensive score is calculated with reference to Table 1, and the comprehensive score is calculated, and the coastal pollution level is determined at the same time, and the result is input into the early warning module, specifically, each indicator in the coastal pollution level judgment table is weighted, and the final total score TP is calculated, where TP is the sum of the scores in the table. If TP≤5, the coastal pollution level is judged to be very clean; if 5<TP≤10, the coastal pollution level is judged to be clean; if 10<TP≤20, the coastal pollution level is judged to be medium clean; if 20<TP≤30, the coastal pollution level is judged to be dirty; if TP>30, the coastal pollution level is judged to be very dirty.
[0051] Table 1 Score calculation of each indicator and coastal pollution level discrimination table
[0052]
[0053]
[0054] S7: In the early warning module, an early warning is generated according to the coastal pollution level and released to the network terminal. As shown in Table 1, the coastal pollution level can be divided into five levels: very clean, clean, medium clean, dirty and very dirty. If the pollution level is very clean, the network terminal displays a dark green prompt; if the pollution level is clean, the network terminal displays a light green prompt; if the pollution level is medium clean, the network terminal displays a yellow prompt; if the pollution level is dirty, the network terminal displays an orange warning; if the pollution level is very dirty, the network terminal displays a red warning and issues an alarm.
[0055] In a second aspect, the present invention also provides a coastal pollution early warning system, including a terrain data acquisition module, a preliminary judgment module, a meteorological data acquisition module, an image data acquisition module, a network identification module, and a comprehensive judgment module;
[0056] Terrain data acquisition module: used to obtain the three-dimensional terrain coordinates and garbage point cloud data of the monitoring area, reconstruct the three-dimensional terrain through Smart3D software, and determine the type of beach in the monitoring area, and then transmit the results to the comprehensive judgment module and early warning module;
[0057] Preliminary judgment module: calculates the actual area of the monitoring area according to the terrain acquisition module, preliminarily judges whether there is garbage according to the garbage distribution, and transmits the judgment result to the comprehensive judgment module;
[0058] Meteorological data acquisition module: used to obtain the wind direction and wind speed in the monitoring area and transmit the data to the comprehensive judgment module;
[0059] Image data acquisition module: used to obtain garbage digital images of the monitoring area and transfer the images to the network recognition module;
[0060] Network recognition module: inputs the digital image passed by the image acquisition module into the pre-trained neural network model, detects whether there is garbage or garbage-like objects in the image, and passes the recognition results to the comprehensive judgment module;
[0061] Comprehensive judgment module: sort out the input data, calculate the coastal pollution level of the monitoring area through the above-mentioned coastal pollution level judgment table, and output the judgment result to the network terminal in real time;
[0062] Early warning module: According to the coastal pollution level judgment result output by the comprehensive judgment module, a real-time early warning is issued. If the judgment level is very clean, the network terminal displays a dark green prompt; if the judgment level is clean, the network terminal displays a light green prompt; if the judgment level is medium clean, the network terminal displays a yellow warning; if the judgment level is dirty, the network terminal displays an orange warning; if the judgment level is very dirty, the network terminal displays a red warning and issues an alarm.
[0063] An embodiment of the present invention also provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements any of the above-mentioned beach garbage identification and classification methods.
[0064] An embodiment of the present invention further provides a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any of the above-mentioned beach garbage identification and classification methods.
[0065] The following is a detailed description of the working principles of the beach garbage identification and classification method and the coastal pollution early warning system.
[0066] like Figure 1-5 As shown in the figure, based on the collected real-time beach images and videos, the trained neural network model is embedded in the early warning system. The garbage in the monitoring area is detected in real time through the neural network model, and the detection results are transmitted to the comprehensive judgment module. By combining the terrain acquisition module, the image data acquisition module, and the network recognition module, the coastal pollution level is comprehensively determined and an early warning is given. The early warning system has a high degree of module integration and can comprehensively analyze the types and quantities of beach garbage. The operation method is simple, and it can effectively identify and classify beach garbage and warn of coastal pollution levels, providing managers with a practical method for beach cleaning management and garbage classification and recycling.
[0067] The above generally describes the present invention in detail, but it is obvious to a person skilled in the art that some modifications or improvements can be made to the present invention. Therefore, modifications or improvements that do not depart from the spirit of the present invention are within the scope of protection of the present invention.
Claims
1. A method for identifying and classifying beach garbage, Features: The following steps are involved: S1: Scan the monitoring area through 3D laser radar to obtain 3D terrain coordinates and garbage point cloud data; S2: Reconstruct the terrain model of the monitoring area through modeling software, calculate the area of the monitoring area, and determine which beach type the area belongs to, and transmit the results to the comprehensive judgment module; S3: Calculate the amount and distribution of garbage in the monitoring area based on the garbage point cloud data, and preliminarily determine whether there is garbage. If there is garbage, the amount of garbage is transmitted to the comprehensive judgment module. If there is no garbage, only the terrain judgment result is transmitted to the comprehensive judgment module. S4: The real-time image data of the monitoring area is obtained through the drone, and input into the trained neural network model to identify whether there are garbage targets in the image, and count the amount of garbage, and transmit the identification and classification results and garbage statistics results to the comprehensive judgment module; S5: In the comprehensive determination module, the clean coast score is calculated according to the input data corresponding to each indicator, and the coastal pollution level is determined at the same time, and the determination result is input into the early warning module; S6: In the early warning module, an early warning is generated according to the coastal pollution level and published to the network terminal.
2. A method for identifying and classifying beach garbage according to claim 1, Features: The modeling software in S2 is Smart3D software.
3. A method for identifying and classifying beach garbage according to claim 1, Features: The point cloud data acquisition software in S3 is a laser radar scanner.
4. A method for identifying and classifying beach garbage according to claim 1, Features: The process of establishing the trained neural network model in S4 includes: Step 1: Use digital cameras and drones to take pictures of beach garbage targets to build a data set; Step 2: Divide the data set into training set and test set in a ratio of 8:2; Step 3: Input the training set into the built neural network model, use back propagation to adjust parameters such as the number of iterations and learning rate, and continue training until the loss function value reaches the minimum, obtain the optimal parameters at this time, and save the optimized model; Step 4: Input the test set into the trained optimization model for testing, and then use the average detection accuracy to evaluate the detection effect of the model.
5. A method for identifying and classifying beach garbage according to claim 1, Features: The coastal pollution level is determined in S5 by using a coastal pollution level discrimination table.
6. A method for identifying and classifying beach garbage according to claim 5, Features: The calculation of the coastal pollution level discrimination table is to assign weights to the indicators in the coastal pollution level discrimination table and calculate the final total score TP.
7. A coastal pollution early warning system, used to implement a beach garbage identification and classification method as described in any one of claims 1 to 6, Features: It includes terrain data acquisition module, preliminary judgment module, meteorological data acquisition module, image data acquisition module, network identification module and comprehensive judgment module; Terrain data acquisition module: used to obtain the three-dimensional terrain coordinate data of the monitoring area, reconstruct the three-dimensional terrain through Smart3D software, and determine the type of beach in the monitoring area, and then transmit the results to the comprehensive judgment module and early warning module; Preliminary judgment module: Preliminary judgment on whether there is garbage based on the beach garbage point cloud data in the terrain acquisition module, and transmit the judgment result to the comprehensive judgment module; Meteorological data acquisition module: used to obtain the wind direction and wind speed in the monitoring area and transmit the data to the comprehensive judgment module; Image data acquisition module: used to obtain digital images of coastal garbage in the monitoring area and transmit the images to the network recognition module; Network recognition module: inputs the digital image passed by the image acquisition module into the pre-trained neural network model to detect whether there is beach garbage in the image, and passes the recognition result to the comprehensive judgment module; Comprehensive judgment module: sort out the input data, calculate the coastal pollution level of the monitoring area through the above-mentioned coastal pollution level judgment table, and output the judgment result to the network terminal in real time; Early warning module: According to the pollution level judgment result output by the comprehensive judgment module, a real-time early warning is issued. If the level is very clean, the network terminal displays a dark green prompt; if the level is clean, the network terminal displays a light green prompt; if the level is medium clean, the network terminal displays a yellow warning; if the level is dirty, the network terminal displays an orange warning; if the level is very dirty, the network terminal displays a red warning, and an alarm is issued.
8. A terminal device, Features: The method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the method for identifying and classifying beach garbage as described in any one of claims 1 to 6 and the coastal pollution early warning system as described in claim 7 are implemented.
9. A computer-readable storage medium, Features: The computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the beach garbage identification and classification method described in any one of claims 1 to 6 and the coastal pollution early warning system described in claim 7.
Citation Information
Cited By
Deep sea sampling area positioning method based on deep learning and multi-modal data
CN120070837A
