Oil spill inspection system based on multispectral and laser radar fusion image recognition

Abstract

The invention relates to the technical field of image recognition, in particular to an oil spill patrol system based on multispectral and laser radar fusion image recognition, which comprises a surface patrol group used for observing the water surface and provided with an image unmanned aerial vehicle cluster carrying a multispectral camera, a plurality of consoles receiving water surface images, and a layered patrol group used for detecting underwater, the system is provided with a radar unmanned aerial vehicle cluster carrying a laser radar, and a central server, and is used for responding to the water surface image and the layered image, determining a corresponding oil spill-like core, and transmitting a corresponding core patrol signal. Oil spill identification of parts below the sea surface is activated in a mode of identifying the oil spill state of the sea surface, modeling is carried out according to the identification result, the sensitivity of oil spill monitoring is effectively improved, meanwhile, energy consumption caused by frequent transfer of an unmanned aerial vehicle carrying a radar is avoided, oil leakage points are identified according to a plurality of detection data, and the oil spill monitoring accuracy is improved. And meanwhile, the detected model is recorded, so that the reliability of oil leakage inspection is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and particularly to an oil spill patrol and survey system based on multi-spectral and lidar fusion image recognition. Background Art

[0002] With the rapid growth of the world economy, the shipping industry has also developed rapidly, and the probability of oil spill accidents has increased accordingly. When an oil spill accident occurs, a large amount of oil spills instantaneously and enters the marine environment, which will cause extremely serious damage to the marine ecological environment.

[0003] In sea areas with severe oil pollution, the probability of red tides will increase. Moreover, with the diffusion and drift of the oil spill, facilities and scenic spots in coastal areas will also be affected, and there is even a possibility of triggering fire and explosion accidents, further exacerbating the losses and impacts of the disasters. Exploring the use of intelligent inspection drones for oil spills on water is an innovative way to enhance the ability to respond to oil spill incidents, reduce the safety risks faced by personnel, and improve the efficiency of oil spill rescue.

[0004] Chinese Patent Authorization Publication No. CN114882371B provides a method for constructing a sea surface oil spill detection model based on full-polarization SAR images, including the following steps: extracting the anisotropic features of full-polarization SAR images, selecting the anisotropic features of oil spill pixels and non-oil spill pixels as training samples, forming matrix A and performing normalization processing to obtain a(t), calculating the pixel weight E from a(t), calculating the pixel feature P from a(t) and the pixel weight E, using the pixel feature P as the training set, and 0 and 1 as the training set labels, where 0 represents non-oil spill pixels and 1 represents oil spill pixels, importing the support vector machine model, and training the model to obtain the oil spill detection model. The invention makes full use of the anisotropic features of full-polarization SAR images and constructs the pixel feature P, improving the accuracy of sea surface oil spill detection in full-polarization SAR images, and having the advantages of being scientific, reasonable, easy to implement, and high in accuracy.

[0005] However, the above method has the following problems: The timeliness of the pictures to be recognized cannot be guaranteed during acquisition, making it difficult to detect oil spill incidents in the early stage of their occurrence. Summary of the Invention

[0006] Therefore, the present invention provides an oil spill patrol and survey system based on multi-spectral and lidar fusion image recognition to overcome the problem in the prior art that the timeliness of the pictures to be recognized cannot be guaranteed during acquisition, resulting in a lag in subsequent responses to oil spill incidents and making it difficult to detect oil spill incidents in the early stage of their occurrence, thereby reducing the oil spill recognition efficiency.

[0007] To achieve the above object, the present invention provides an oil spill patrol and survey system based on multi - spectral and lidar fusion image recognition, including: A surface inspection group for observing the water surface, which is provided with an image drone cluster carrying a multi - spectral camera. Each cluster collects water surface images using visible light and / or infrared light according to a collection period; A number of consoles for receiving the water surface images, which are used to determine the corresponding oil - spill - like areas according to the water surface images; A stratified inspection group for detecting underwater, which is provided with a lidar drone cluster carrying lidar, and is used to collect stratified images of the oil - spill - like areas using the lidar; A central server, which is connected to each console, and is used to determine the corresponding oil - spill core according to the water surface images and the stratified images, and send out corresponding core patrol and survey signals; Among them, the console responds to the core patrol and survey signals, controls the corresponding lidar drone clusters to take pictures of various oil - spill - like areas to generate corresponding oil - spill images; The collection period corresponds to the longest patrol and survey duration of each drone composing the image drone cluster; The central server collects the water surface images, the stratified images and the oil - spill images and forms an oil - spill monitoring database.

[0008] Further, the console is composed of a drifting console and a shore - based console. Among them, any console is provided with a corresponding patrol and survey range, and the patrol and survey ranges of each console cover the sea area to be patrolled in any collection period.

[0009] Further, the surface inspection group includes a number of hangars corresponding to each console, and a number of image drone clusters set based on each hangar; Among them, for a single console, any drone in the corresponding drone cluster starts from the hangar corresponding to this console during a single patrol and survey, and returns to the hangar corresponding to this console when completing the corresponding patrol and survey route; The patrol and survey route of the drone is related to the patrol and survey range of the console corresponding to the drone, and the image coverage range of its single - time patrol and survey is not less than a preset ratio of the patrol and survey range; The preset ratio is related to the viewport of the drone.

[0010] Further, the lidar drone group includes: A flight station set on the shore - based console, and a lidar drone cluster composed of a number of lidar drones; Among them, the number of lidar drones is equal to the number of flight stations. When a single lidar drone completes the shooting of the stratified image, this lidar drone enters the flight station with the shortest relative distance to it.

[0011] Furthermore, the central server is provided with an oil spill color difference value and an oil spill area threshold. When any adjacent area of the water surface image within any patrol range reaches the oil spill color difference value, the central server issues a color difference warning; When the water surface area of the water surface image within any acquisition period reaching the oil spill color difference value is not less than the oil spill area threshold, the central server issues a stratification warning, and determines the corresponding oil spill-like area and the console corresponding to the oil spill-like area; Among them, the oil spill color difference value is the gray level difference value between the sea surface color and the oil spill color, which is related to the lighting conditions; The oil spill area threshold is related to the maximum oil spill range of a single oil spill event within a single acquisition period.

[0012] Furthermore, the console corresponding to the oil spill-like area responds to the color difference warning, and sets a number of adjacent consoles as extended consoles; Each extended console responds to the color difference warning and controls the corresponding image unmanned cluster to collect the corresponding patrol range; The central server responds to the color difference warning and extends the acquisition period until each extended console completes the corresponding water surface image acquisition.

[0013] Furthermore, the central server responds to the stratification warning, sets at least one shore-based console closest to the center of the oil spill-like area corresponding to the stratification warning as the active console, and issues a stratification image drawing instruction with at least one shore-based console closest to the edge of the oil spill-like area as the slave console.

[0014] Furthermore, the active console and the slave console respond to the stratification image drawing instruction and respectively control the corresponding radar drones to detect the stratification image of the oil spill-like area along a preset route; Among them, the preset route is related to the contour of the oil spill-like area and the corresponding ocean current direction.

[0015] Furthermore, the central server determines a number of oil spill cores based on the stratification image and issues corresponding core patrol signals; The stratification patrol group responds to the core patrol signal and detects the oil spill core along the preset route; Among them, the core patrol signal is sent to each shore-based console; The oil spill core is at the bottom layer of the stratification image.

[0016] Furthermore, when the central server obtains the corresponding stratification image of the oil spill core, it generates a corresponding oil spill image; Among them, the oil spill image forms a corresponding longitudinal mapping from the layered image and a corresponding water surface mapping from the water surface image.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows. By setting the unmanned aerial vehicle (UAV) as a surface inspection group and a layered inspection group, and setting a number of consoles and a central server, the oil spill identification below the sea surface is activated by identifying the oil spill state on the sea surface, and a model is built based on the identification result. While effectively improving the sensitivity of oil spill monitoring, it avoids the energy consumption caused by frequently mobilizing the UAV equipped with a radar, and can identify the oil spill point according to a number of detection data, and record the completed detection model at the same time, thus effectively improving the reliability of oil spill patrol and measurement.

[0018] Furthermore, by setting up a drifting console and a shore-based console to conduct inspections on the sea surface respectively, it avoids the situation that the oil spill phenomenon cannot be observed due to the oil spill position exceeding the patrol and measurement radius of the UAV, effectively improving the extensibility of the oil spill patrol and measurement system and further enhancing the reliability of oil spill patrol and measurement.

[0019] Furthermore, by setting the corresponding patrol and measurement range according to the viewport of the UAV, while improving the patrol and measurement range of the UAV, it avoids the problem that the sea surface cannot be observed due to the mismatch between the shooting viewport size of the UAV and the patrol and measurement range, thus further enhancing the reliability of oil spill patrol and measurement.

[0020] Furthermore, by observing the color difference of the sea surface and presetting an oil spill area threshold to characterize the degree of oil spill diffusion caused by seawater flow, while effectively improving the timeliness of oil spill patrol and measurement, it avoids the situation that when a large oil spill area is found during detection, it is impossible to accurately determine whether the stage of the oil spill event belongs to the early stage or the diffusion stage, thus further enhancing the reliability of oil spill patrol and measurement.

[0021] Furthermore, by using the layered image as the longitudinal model benchmark and the water surface image as the horizontal model benchmark to build a model for the oil spill event, and storing the completed oil spill model, using this model, a reasonable oil spill source area can be found through model analysis, which can provide model support for subsequent oil spill events, thus further enhancing the reliability of oil spill patrol and measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic structural diagram of the oil spill patrol and measurement based on the fusion image of multi-spectral and lidar of the present invention; Figure 2 It is a schematic connection diagram of the shore-based console of the oil spill patrol and measurement system in the embodiment of the present invention; Figure 3 It is a schematic connection diagram of the drifting console of the oil spill patrol and measurement system in the embodiment of the present invention; Figure 4 Schematic diagram of communication between the oil spill survey drone and the ground base station in the embodiment of the present invention; Figure 5 Oil spill model in the embodiment of the present invention; In the figure: 1, water surface image; 11, water surface oil spill range; 2, layered image; 21, layered oil spill range; 3, oil spill target layer; 31, speculated oil spill core; 4, water surface depth; 41, first layered depth; 42, second layered depth; 43, third layered depth; 44, oil spill core layered depth. Specific embodiments

[0023] In order to make the objectives and advantages of the present invention clearer and more understandable, the present invention will be further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0024] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present invention and do not limit the protection scope of the present invention.

[0025] It should be noted that in the description of the present invention, the terms indicating directions or positional relationships such as "upper", "lower", "left", "right", "inner", "outer", etc. are based on the directions or positional relationships shown in the drawings. This is only for convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention.

[0026] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installation", "connection", and "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0027] Please refer to Figure 1 As shown, it is a schematic structural diagram of the oil spill survey based on the fusion image recognition of multi-spectral and lidar of the present invention, including: A surface inspection group for observing the water surface, which is equipped with an image drone cluster carrying multi-spectral cameras. Each cluster collects water surface images using visible light and / or infrared light according to the collection period; Several consoles for receiving water surface images, which are used to determine the corresponding oil spill-like areas according to the water surface images; A layered inspection group for detecting underwater, which is equipped with a cluster of radar drones carrying lidar, for collecting layered images of oil spill-like areas using radar; A central server, which is connected to each console, for determining corresponding oil spill-like cores according to the water surface images and layered images and sending corresponding core inspection signals; Among them, the console responds to the core inspection signal and controls the corresponding radar drone clusters to take pictures of various oil spill areas to generate corresponding oil spill images; The acquisition period corresponds to the longest inspection duration corresponding to each drone that makes up the image drone cluster; The central server collects water surface images, layered images, and oil spill images and forms an oil spill monitoring database.

[0028] By setting the drones as a surface inspection group and a layered inspection group, and setting several consoles and a central server, the oil spill identification of the part below the sea surface is activated by identifying the oil spill state of the sea surface, and modeling is carried out according to the identification results. While effectively improving the sensitivity of oil spill monitoring, it avoids the energy consumption caused by frequently mobilizing drones equipped with radar, and can identify oil leakage points according to several detection data, and record the completed detection model at the same time, thus effectively improving the reliability of oil leakage inspection.

[0029] Please refer to Figure 2 and Figure 3 As shown, they are respectively the connection schematic diagram of the shore-based console of the oil spill inspection system according to the embodiment of the present invention and the connection schematic diagram of the drifting console of the oil spill inspection system according to the embodiment of the present invention. Among them, the console is composed of a drifting console and a shore-based console. Among them, any console is provided with a corresponding inspection range, and the inspection ranges of each console cover the sea area to be inspected in any acquisition period.

[0030] By setting the drifting console and the shore-based console to conduct inspections on the sea surface respectively, it avoids the situation that the oil spill phenomenon cannot be observed due to the oil spill position exceeding the inspection radius of the drone, effectively improving the extensibility of the oil spill inspection system and further improving the reliability of the oil spill inspection.

[0031] In particular, the surface inspection group includes several hangars corresponding to each console, and several image drone clusters based on each hangar; Among them, for a single console, any drone in the corresponding drone cluster starts from the hangar corresponding to this console during a single inspection and returns to the hangar corresponding to this console when completing the corresponding inspection route; The inspection route of the drone is related to the inspection range of the console corresponding to this drone, and the image coverage range of its single inspection is not less than a preset ratio of the inspection range; The preset ratio is related to the viewport of the drone.

[0032] By setting the corresponding survey range according to the viewport of the drone, while improving the survey range of the drone, it avoids the problem that the sea surface cannot be observed due to the mismatch between the shooting viewport size of the drone and the survey range, thereby further improving the reliability of oil spill survey.

[0033] In particular, the radar drone group includes: A flight station set on the shore-based console, and a radar drone cluster composed of several radar drones; Among them, the number of radar drones is equal to the number of flight stations. When a single radar drone completes the shooting of the layered image, the radar drone enters the flight station with the shortest relative distance to it.

[0034] Specifically, the central server is provided with an oil spill color difference value and an oil spill area threshold. When any adjacent area of the water surface image in any survey range reaches the oil spill color difference value, the central server issues a color difference warning; When the water surface area of the water surface image in any acquisition period reaching the oil spill color difference value is not less than the oil spill area threshold, the central server issues a layered warning, and determines the corresponding oil spill-like area and the console corresponding to the oil spill-like area; Among them, the oil spill color difference value is the gray difference value between the sea surface color and the oil spill color, which is related to the lighting conditions; The oil spill area threshold is related to the maximum oil spill range of a single oil spill event in a single acquisition period.

[0035] In one embodiment, under strong daylight lighting conditions, the oil spill color difference value set by the central server is 30, and the oil spill area threshold is 500 square meters. During a survey, the gray difference value between the sea surface color and the oil spill color in a certain area shown in the water surface image reaches 35, exceeding the set oil spill color difference value.

[0036] Warning situation: Color difference warning: Since the gray difference value 35 is greater than the oil spill color difference value 30, the central server issues a color difference warning.

[0037] Layered warning: After further analysis, it is found that the water surface area reaching the oil spill color difference value is 600 square meters, exceeding the oil spill area threshold of 500 square meters. The central server issues a layered warning, determines that this area is an oil spill-like area, and notifies the corresponding console for further processing.

[0038] In one embodiment, under weak light conditions at night, the central server sets the oil spill color difference value to 20 and the oil spill area threshold to 300 square meters. During a patrol survey, the grayscale difference value between the sea surface color and the oil spill color in a certain area shown in the water surface image is 18, which does not reach the set oil spill color difference value.

[0039] Early warning situation: Color difference early warning: Since the grayscale difference value of 18 is less than the oil spill color difference value of 20, the central server does not issue a color difference early warning.

[0040] Stratification early warning: Since no color difference early warning is issued, the central server will not conduct further stratification early warning analysis.

[0041] In one embodiment, under moderately cloudy weather conditions with moderate lighting, the central server sets the oil spill color difference value to 25 and the oil spill area threshold to 400 square meters. During a patrol survey, the grayscale difference value between the sea surface color and the oil spill color in a certain area shown in the water surface image is 28, which exceeds the set oil spill color difference value.

[0042] Early warning situation: Color difference early warning: Since the grayscale difference value of 28 is greater than the oil spill color difference value of 25, the central server issues a color difference early warning.

[0043] Stratification early warning: After further analysis, it is found that the water surface area reaching the oil spill color difference value is 450 square meters, which exceeds the oil spill area threshold of 400 square meters. The central server issues a stratification early warning, determines that this area is an oil spill-like area, and notifies the corresponding console for further processing.

[0044] It can be understood that the above embodiments only represent one possible situation. In actual applications, corresponding settings should be made according to the actual situation of the ocean surface.

[0045] Please refer to Figure 4 as shown, which is a communication schematic diagram of an oil spill patrol unmanned aerial vehicle and a ground base station according to an embodiment of the present invention, including: Drone platform carrying: Drone platform carrying generally refers to the body of the drone and its support structure, including components such as wings, fuselage, landing gear, and their connection, support, and loading facilities. It not only has to support the flight of the drone but also carry various payloads required for performing tasks, such as sensors, cameras, communication equipment, etc.

[0046] Multispectral camera: A multispectral camera is an advanced imaging device that can simultaneously obtain spectral characteristics and spatial image information. A multispectral camera can not only capture images in the visible light band but also sense spectral ranges invisible to the human eye, such as near-infrared and short-wave infrared. It can detect oil film areas more accurately.

[0047] Micro LiDAR: It consists of a laser emitter, an optical system, a scanning system, a receiver, and a signal processing unit. The working principle of LiDAR is based on Time of Flight (TOF) technology. The LiDAR system emits laser pulses and records the time difference between the emission of the laser pulses and the reception of the reflected pulses. By knowing the speed of light, the distance that the laser pulse travels to and from the target can be calculated.

[0048] Thermal infrared video capture card: Usually equipped with a dedicated thermal infrared sensor, it can convert thermal radiation into electrical signals.

[0049] Visible light video capture card: It can convert the analog video signal output by the camera into a digital signal so that the computer's operating system and software can process it.

[0050] Radar data acquisition card: It can convert the analog or digital signals generated by the radar system into digital data that can be processed by a computer.

[0051] Radar data processor: The radar data processor is a key component in the radar system. It is responsible for processing the raw signals received by the radar to extract important information about the target. It can reduce the influence of sea waves on oil film detection.

[0052] Oil film thickness inversion module: It calculates the thickness of the oil film by detecting the influence of the oil film on specific signals.

[0053] Thermal infrared image, visible light image, and radar image fuser: It integrates the image data from different sensors for subsequent analysis and processing.

[0054] Oil spill detection and analysis card: It is used to detect and analyze the oil spill information in the fused images.

[0055] Oil spill detection alarm: The oil spill detection alarm is equipped with an electronic display screen and is installed at the server end of the command center. Its function is to display the location of the suspected oil spill on the electronic nautical chart in a flashing manner and to prompt the area of the suspected oil spill with a sound alarm. After the oil spill detection and analysis card in the ground console detects an oil spill on the sea surface, it sends the suspected oil spill analysis results (including the location and area of the suspected oil spill), the thermal infrared, visible light, and radar fused images, the thermal infrared image, the visible light image, and the radar image to the server end of the command center.

[0056] Oil spill monitoring database: The oil spill monitoring database is deployed at the server end of the command center and is used to store the suspected oil spill analysis results (including the location and area of the suspected oil spill), the thermal infrared, visible light, and radar fused images, the thermal infrared image, the visible light image, the radar image, and the alarm time information.

[0057] Specifically, the console corresponding to the oil spill-like area responds to the color difference warning and sets several adjacent consoles as extended consoles; Each extended console responds to the color difference warning and controls the corresponding image unmanned cluster to collect the corresponding patrol range; The central server responds to the color difference warning and extends the collection period until each extended console completes the collection of the corresponding water surface image.

[0058] By observing the color difference of the sea surface and presetting an oil spill area threshold to characterize the degree of oil spill diffusion caused by seawater flow, while effectively improving the timeliness of oil spill patrol, it avoids the inability to accurately determine whether the stage of the oil spill event is in the early stage or the diffusion stage when a large oil spill area is found during detection, thereby further enhancing the reliability of oil spill patrol.

[0059] Specifically, the central server responds to the hierarchical warning and sets at least one shore-based console closest to the center of the oil spill-like area corresponding to the hierarchical warning as the active console, and, Sets at least one shore-based console closest to the edge of the oil spill-like area as the slave console to issue a hierarchical image drawing instruction.

[0060] Specifically, the active console and the slave console respond to the hierarchical image drawing instruction and respectively control the corresponding radar drones to detect the hierarchical image of the oil spill-like area along a preset route; Among them, the preset route is related to the contour of the oil spill-like area and the corresponding ocean current direction.

[0061] Specifically, the central server determines several oil spill cores based on the hierarchical image and issues corresponding core patrol signals; The hierarchical inspection group responds to the core patrol signal and detects the oil spill core along the preset route; Among them, the core patrol signal is sent to each shore-based console; The oil spill core is at the bottom layer of the hierarchical image.

[0062] Specifically, when the central server obtains the corresponding hierarchical image of the oil spill core, it generates a corresponding oil spill image; Among them, the oil spill image forms a corresponding longitudinal mapping from the hierarchical image and a corresponding water surface mapping from the water surface image.

[0063] By using the hierarchical image as the longitudinal model benchmark and the water surface image as the horizontal model benchmark to model the oil spill event and storing the completed oil spill model, using this model, a reasonable oil spill source area can be found through model analysis, which can provide model support for subsequent oil spill events, thereby further enhancing the reliability of oil spill patrol.

[0064] Please refer to Figure 5As shown, it is a schematic diagram of the oil spill model according to an embodiment of the present invention. In the figure, when the oil spill range 11 in the water surface image 1 exceeds the set range, the central server divides this range into a rectangular area including several consoles, and controls the radar UAV group corresponding to the hierarchical inspection group to detect the images at each depth; At this time, based on the water surface depth 4, the radar UAV divides the first hierarchical depth 41, the second hierarchical depth 42, and the third hierarchical depth 43 corresponding to different radar frequencies, and respectively detects the hierarchical oil spill ranges 21 corresponding to each hierarchical image 2, and transmits each hierarchical image 2 to the central server; At this time, the central server infers several smooth fitting curves based on the hierarchical oil spill ranges 21 corresponding to each hierarchical image 2 and the oil spill range 11 corresponding to the water surface image 1, maps the oil spill range 11 on the water surface and each hierarchical oil spill range 21 vertically to form the corresponding inferred oil spill core 31 and the corresponding oil spill target layer 3, and forms the modeling of this oil spill event.

[0065] So far, the technical solution of the present invention has been described in combination with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0066] The above are only the preferred embodiments of the present invention and are not used to limit the present invention; for those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An oil spill inspection system based on multi-spectral and laser radar fusion image recognition, characterized in that: include: A surface inspection team for observing the water surface, which is equipped with a cluster of imaging drones carrying multi-spectral cameras, each cluster uses visible light and / or infrared light to collect water surface images according to a collection cycle; A plurality of control consoles receiving the water surface image are used to determine the corresponding oil spill-like area according to the water surface image; An underwater stratified inspection team is provided with a radar drone cluster carrying a laser radar to collect stratified images of the oil spill area using radar; A central server connected to each console, for responding to the water surface image and the layered image to determine the corresponding oil spill-like core, and to send a corresponding core patrol signal; The control console responds to the core patrol signal and controls the corresponding radar drone clusters to photograph various oil spill areas to generate corresponding oil spill images; The acquisition period corresponds to the longest patrol duration of each drone constituting the image drone cluster; The central server collects the water surface image, the layered image and the oil spill image and forms an oil spill monitoring database.

2. The oil spill patrol system based on multi-spectral and laser radar fusion image recognition according to claim 1 is characterized in that: The control console is composed of a drifting control console and a shore-based control console, wherein any control console is provided with a corresponding patrol range, and the patrol range of each control console covers the sea surface to be patrolled in any collection cycle.

3. The oil spill inspection system based on multi-spectral and laser radar fusion image recognition according to claim 2 is characterized in that: The surface inspection group includes a number of machine nests corresponding to each control console, and a number of image drone clusters based on each machine nest; Among them, for a single control station, any drone of the corresponding drone cluster will start from the machine nest corresponding to the control station in a single patrol, and return to the machine nest corresponding to the control station when completing the corresponding patrol route; The inspection route of the drone is related to the inspection range of the control console corresponding to the drone, and the image coverage of a single inspection is not less than a preset proportion of the inspection range; The preset ratio is related to the spatial resolution of the drone.

4. The oil spill patrol system based on multi-spectral and laser radar fusion image recognition according to claim 2 is characterized in that: The radar drone group includes: A flight station disposed at the shore-based control station, and a radar drone cluster consisting of a number of radar drones; The number of the radar drones is equal to the number of the flight stations. When a single radar drone completes the shooting of the layered image, the radar drone enters the flight station with the shortest relative distance thereto.

5. The oil spill patrol system based on multi-spectral and laser radar fusion image recognition according to any one of claims 3 or 4 is characterized in that: The central server is provided with an oil spill color difference value and an oil spill area threshold. When any adjacent area of ​​a water surface image within any patrol range reaches the oil spill color difference value, the central server issues a color difference warning; When the water surface area of ​​the water surface image in any acquisition period reaches the oil spill color difference value and is not less than the oil spill area threshold, the central server issues a layered warning and determines the corresponding oil spill-like area and the control console of the corresponding oil spill-like area; The oil spill color difference value is the grayscale difference between the sea surface color and the oil spill color, which is related to the lighting conditions; The oil spill area threshold is related to the maximum oil spill range of a single oil spill incident within a single collection cycle.

6. The oil spill patrol system based on multi-spectral and laser radar fusion image recognition according to claim 5 is characterized in that: The control console corresponding to the oil spill-like area responds to the color difference warning and sets several adjacent control consoles as extended control consoles; Each extended control console responds to the color difference warning and controls the corresponding image unmanned cluster to collect the corresponding patrol range; In response to the color difference warning, the central server extends the acquisition period until each extended console completes the corresponding water surface image acquisition.

7. The oil spill inspection system based on multi-spectral and laser radar fusion image recognition according to claim 5 is characterized in that: The central server responds to the layered warning, sets at least one shore-based control station closest to the center of the oil spill-like area corresponding to the layered warning as an active control station, and At least one shore-based control station closest to the edge of the oil spill-like area is used as a slave control station to issue a layered image drawing instruction.

8. The oil spill inspection system based on multi-spectral and laser radar fusion image recognition according to claim 7 is characterized in that: The active control console and the passive control console respond to the layered image drawing instruction, and respectively control the corresponding radar drones to detect the layered image of the oil spill area along a preset route; The preset route is related to the outline of the oil spill-like area and the corresponding ocean current direction.

9. The oil spill patrol system based on multi-spectral and laser radar fusion image recognition according to claim 8 is characterized in that: The central server determines several types of oil spill cores according to the layered image and issues corresponding core patrol signals; The layered inspection team responds to the core inspection signal and detects the oil spill core along the preset route; Wherein, the core patrol signal is sent to each shore-based control console; The oil spill core is at the bottom layer of the layered image.

10. The oil spill patrol system based on multi-spectral and laser radar fusion image recognition according to claim 9 is characterized in that: The central server generates a corresponding oil spill image when acquiring the corresponding layered image of the oil spill core; The oil spill image is formed by the layered image to form a corresponding longitudinal mapping, and the water surface image is formed by the water surface image to form a corresponding water surface mapping.

Images