Ecological wetland data monitoring method and system based on hyperspectral imaging system

Abstract

The application relates to the technical field of data monitoring, and particularly discloses an ecological wetland data monitoring method and system based on a hyperspectral imaging system, which comprises the following steps: acquiring a map of a wetland area, selecting a UAV according to the map, and synchronously determining a movement path of the UAV; receiving wetland images with position labels and time labels fed back by the UAV at a fixed time, identifying the wetland images, and extracting wetland data; identifying the wetland data, determining a movement speed of the UAV, generating a movement instruction according to the movement speed, and sending the movement instruction to the UAV. The application acquires the collection results of the UAV at a fixed time, identifies the collection results, obtains the wetland data, accelerates the movement speed of the UAV when the wetland data is in a standard state, and improves the endurance in a way. In addition, the UAV does not need to feed back the collection results in real time, consumption of transmission resources is reduced, and the endurance is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data monitoring, and particularly relates to an ecological wetland data monitoring method and system based on a hyperspectral imaging system. BACKGROUND

[0002] The application of a light and small airborne hyperspectral imaging system in wetland data acquisition mainly helps researchers to deeply analyze environmental changes, species distribution and hydrological characteristics of a wetland ecological system through the advantages of high spatial resolution and high spectral resolution of the system. The existing light and small airborne hyperspectral imaging system is generally carried on a drone, and the drone acquires ecological wetland data. In this process, the endurance of the drone is very important. The drone needs to be charged regularly. When the number of drones is large, the charging process is very cumbersome. Even if the endurance is improved to some extent, great benefits can be brought. How to improve the endurance is a technical problem to be solved by the technical scheme of the present application. SUMMARY

[0003] The present application aims to provide an ecological wetland data monitoring method and system based on a hyperspectral imaging system to solve the problems in the background art.

[0004] To achieve the above object, the present application provides the following technical scheme.

[0005] An ecological wetland data monitoring method based on a hyperspectral imaging system, the method comprising:

[0006] acquiring a map of a wetland area, selecting a drone according to the map, and synchronously determining a motion path of the drone; the drone is built-in with a hyperspectral imaging system;

[0007] receiving wetland images with position labels and time labels fed back by the drone regularly, identifying the wetland images, and extracting wetland data;

[0008] identifying the wetland data, determining a motion speed of the drone, generating a motion instruction according to the motion speed, and sending the motion instruction to the drone;

[0009] acquiring an actual speed of the drone by a positioning device built-in in the drone, comparing the actual speed with the motion speed to mark a collection time, reading an image of the marked collection time when the drone returns;

[0010] In the motion process of the drone, the wetland images collected are identified in real time locally, and the received motion instruction is adjusted according to the local identification result.

[0011] As a further scheme of the present application, the step of acquiring a map of a wetland area, selecting a drone according to the map, and synchronously determining a motion path of the drone comprises:

[0012] acquire and display a map containing a wet area, receive a user inputted area boundary, determine a wet area;

[0013] receive a user inputted working parameter of a UAV, determine a cutting width according to the working parameter; the working parameter includes a working height and a working wide angle;

[0014] uniformly set sampling points on the area boundary, sequentially query the farthest sampling point of each sampling point, and acquire the distance thereof;

[0015] select and connect two sampling points with the largest distance as a direction line;

[0016] based on the direction line and the cutting width array, calculate the intersection of the direction line after the array and the wet area, sequentially connect the head and tail of the intersection, and obtain a motion path.

[0017] As a further scheme of the present application, the step of receiving a wet image containing a position label and a time label feedback by the UAV, and identifying the wet image to extract wetland data includes:

[0018] sending the motion path to at least one UAV, and receiving a wet image containing a position label and a time label feedback by the UAV;

[0019] determining the wetland image of different wave bands at the latest time of each position according to the position label and the time label;

[0020] for each wave band, splicing all the wetland images to obtain the latest image of the entire wet area;

[0021] identifying the latest image of each wave band, recording the identification time, and obtaining wetland data containing a time label; the wetland data is an array, and each element corresponds to a parameter.

[0022] As a further scheme of the present application, the step of identifying the wetland data, determining the motion speed of the UAV, generating a motion instruction according to the motion speed, and sending the motion instruction to the UAV includes:

[0023] arranging the wetland data containing a time label in time sequence to obtain a data matrix;

[0024] cutting the data matrix according to a preset time span, comparing the cut data matrix with a preset standard data group, and determining the motion speed of the UAV according to the comparison result;

[0025] generating a motion instruction according to the motion speed, and sending the motion instruction to the UAV;

[0026] wherein, the determination process of the motion speed is:

[0027] Where, Indicates the speed of movement, and Represent the intercepted data matrix and standard data set respectively, express and The number of intersection elements; the process of determining the intersection element is: when the difference between the data in the intercepted data matrix and the corresponding data in the standard data group is less than a preset threshold, it is regarded as an intersection element; Indicates the total number of elements in the intercepted data matrix; is the preset correction factor.

[0028] As a further solution of the present invention, the steps of obtaining the actual speed of the drone based on the positioning device built into the drone, comparing the actual speed with the movement speed to mark the acquisition time, and reading the image at the marked acquisition time when the drone returns include:

[0029] Obtain the drone's location in real time based on the built-in positioning device of the drone;

[0030] Calculate the actual speed of the drone based on the position;

[0031] When the actual speed is less than the movement speed, mark the corresponding moment;

[0032] When the drone returns, the image at the marked acquisition moment is read.

[0033] As a further solution of the present invention: the working process of the drone includes:

[0034] Receive movement instructions;

[0035] Collect wetland images in real time and locate abnormal features and their levels in the wetland images based on a preset convolutional recognition model;

[0036] Determine the risk level of wetland images based on the level and number of abnormal features;

[0037] The motion command is adjusted according to the risk level; wherein, the greater the risk level, the smaller the speed corresponding to the adjusted motion command.

[0038] The technical solution of the present invention also provides an ecological wetland data monitoring system based on a hyperspectral imaging system, the system comprising:

[0039] A motion path determination module is used to obtain a map of the wetland area, select a drone based on the map, and simultaneously determine the motion path of the drone; the drone has a built-in hyperspectral imaging system;

[0040] The wetland data extraction module is configured to receive the wetland images with position tags and time tags fed back by the unmanned aerial vehicle in a time manner, identify the wetland images, and extract wetland data.

[0041] The wetland data identification module is configured to identify the wetland data, determine the moving speed of the unmanned aerial vehicle, generate a moving instruction according to the moving speed, and send the moving instruction to the unmanned aerial vehicle.

[0042] The marker reading module is configured to acquire an actual speed of the unmanned aerial vehicle according to a positioning device built in the unmanned aerial vehicle, compare the actual speed with the moving speed, read the image of the collection time of the marker when the unmanned aerial vehicle returns, and read the image of the collection time of the marker.

[0043] In the process of moving, the unmanned aerial vehicle performs real-time local identification on the collected wetland images, and adjusts the received moving instruction according to the local identification result.

[0044] As a further scheme of the present application, the moving path determination module comprises:

[0045] The region determination unit is configured to acquire and display a map containing a wetland region, receive a region boundary input by a user, and determine the wetland region.

[0046] The width determination unit is configured to receive a working parameter of the unmanned aerial vehicle input by the user, and determine a segmentation width according to the working parameter; the working parameter comprises a working height and a working wide angle.

[0047] The distance acquisition unit is configured to uniformly set sampling points on the region boundary, sequentially query the farthest sampling point of each sampling point, and acquire the distance.

[0048] The direction line determination unit is configured to select and connect the two sampling points with the largest distance as a direction line.

[0049] The connection unit is configured to calculate the intersection of the direction line after arraying and the wetland region based on the arrayed direction line and the segmentation width, sequentially connect the intersection from the head to the tail, and obtain the moving path.

[0050] As a further scheme of the present application, the wetland data extraction module comprises:

[0051] The image receiving unit is configured to send the moving path to at least one unmanned aerial vehicle, and receive the wetland images with position tags and time tags fed back by the unmanned aerial vehicle.

[0052] The image selection unit is configured to determine the wetland images of different wave bands at the latest time of each position according to the position tags and the time tags.

[0053] The image splicing unit is configured to splice all the wetland images for each wave band to obtain the latest image of the entire wetland region.

[0054] a recording unit, configured to identify the latest image of each waveband, record the identification time, and obtain wetland data with time labels; the wetland data is an array, and each element corresponds to a parameter.

[0055] As a further scheme of the present application, the wetland data identification module comprises:

[0056] a data arrangement unit, configured to arrange the wetland data with time labels according to time sequence, and obtain a data matrix;

[0057] a comparison execution unit, configured to intercept the data matrix according to a preset time span, compare the intercepted data matrix with a preset standard data set, and determine the movement speed of the unmanned aerial vehicle according to the comparison result;

[0058] an instruction generation and sending unit, configured to generate a movement instruction according to the movement speed, and send the movement instruction to the unmanned aerial vehicle;

[0059] wherein the determination process of the movement speed is as follows:

[0060] ; in the formula, the movement speed is represented by V, and the intercepted data matrix and the standard data set are represented by X and Y respectively, represents and the number of intersection elements; the determination process of the intersection elements is as follows: when the difference between the data in the intercepted data matrix and the corresponding data in the standard data set is less than a preset threshold value, the data is regarded as an intersection element; represents the total number of elements in the intercepted data matrix; is a preset correction coefficient.

[0061] Compared with the prior art, the present application has the beneficial effects that: the present application obtains the collection results of the unmanned aerial vehicle at regular time intervals, identifies the collection results, and obtains wetland data; when the wetland data is in a standard state, the movement speed of the unmanned aerial vehicle is accelerated, and the endurance is improved indirectly; in addition, the unmanned aerial vehicle does not need to feed back the collection results in real time, the consumption of transmission resources is reduced, and the endurance is improved. BRIEF DESCRIPTION OF DRAWINGS

[0062] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only some embodiments of the present application.

[0063] Figure 1 is a flowchart of the ecological wetland data monitoring method based on the hyperspectral imaging system.

[0064] Figure 2 The first sub-flow block diagram of the ecological wetland data monitoring method based on the hyperspectral imaging system.

[0065] Figure 3 The second sub-flow block diagram of the ecological wetland data monitoring method based on the hyperspectral imaging system.

[0066] Figure 4 The third sub-flow block diagram of the ecological wetland data monitoring method based on the hyperspectral imaging system.

[0067] Figure 5 The fourth sub-flow block diagram of the ecological wetland data monitoring method based on the hyperspectral imaging system.

[0068] Figure 6 The composition structure block diagram of the ecological wetland data monitoring system based on the hyperspectral imaging system. DETAILED DESCRIPTION

[0069] In order to make the technical problems to be solved by the present application, technical solutions and beneficial effects clearer, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0070] Figure 1 The flow block diagram of the ecological wetland data monitoring method based on the hyperspectral imaging system, in the embodiment of the present application, a kind of ecological wetland data monitoring method based on the hyperspectral imaging system, the method comprises:

[0071] Step S100: obtain the map of wetland area, select unmanned aerial vehicle according to the map, the motion path of unmanned aerial vehicle is determined synchronously;The hyperspectral imaging system is built-in in the unmanned aerial vehicle;

[0072] Wetland area is the area that needs to be monitored, obtain the map of wetland area, select unmanned aerial vehicle according to the map, determine the motion path of unmanned aerial vehicle, so that it can carry out data acquisition to wetland area, and data acquisition process is completed by hyperspectral imaging system built-in in unmanned aerial vehicle, and the hyperspectral imaging system captures the spectral characteristics of ground object through multiple continuous wave bands, and the information of these wave bands can reflect the change of multiple hydrological parameters, especially in wetland and water body monitoring aspect has important significance.Each wave band corresponds to different spectral characteristics, and can reflect different physical and chemical characteristics of water body;Some key hydrological parameters are as follows:

[0073] 1. water turbidity (Turbidity):

[0074] The turbidity of a water body is closely related to the concentration of suspended particulate matter in the water. Different wavelengths of spectral information can reflect the content of suspended particles (such as sediment, algae, plankton, etc.) in the water. Generally speaking, the short-wave infrared (SWIR) band (1000nm-2500nm) is very sensitive to water turbidity, especially between 1500nm-1750nm and 1900nm-2100nm, the concentration of suspended solids in water will cause significant changes in reflectivity.

[0075] 2. Water quality pollutants (such as dissolved organic matter, heavy metals):

[0076] Dissolved substances in water (such as dissolved oxygen, organic matter, heavy metals, etc.) have specific absorption and reflection characteristics. For example, dissolved organic matter in water (such as humic acid, etc.) usually exhibits strong absorption characteristics in the ultraviolet band (200nm-400nm) and the visible to near-infrared band (400nm-1000nm). Hyperspectral systems can identify these characteristics to infer the pollution level of the water body.

[0077] 3. Water temperature:

[0078] Changes in water temperature will affect its spectral reflectance, especially the short-wave infrared (SWIR) band (1300nm-2500nm) is very sensitive to changes in water temperature. In this band, the reflectivity of water will change with the increase of temperature. Hyperspectral imaging systems can infer the surface temperature of the water body by analyzing the reflectivity of different bands combined with thermal infrared data.

[0079] 4. Algal concentration (algae bloom monitoring):

[0080] The growth and reproduction of algae will affect the spectral characteristics of the water body, especially in the near-infrared (NIR) band (700nm-1300nm) and the red edge band (700nm-750nm) have significant reflection characteristics. Algae in water (especially planktonic algae) have strong reflectivity in these bands. Hyperspectral imaging can help monitor the concentration of algae in water, especially when algae blooms occur (such as blue-green algae, green algae, etc.), the change of reflectivity is particularly significant.

[0081] 5. Dissolved oxygen concentration in water:

[0082] The concentration of dissolved oxygen in water will affect its spectral characteristics, especially in certain specific bands (such as the near-infrared band) the concentration of dissolved oxygen in water will have a certain impact on the absorption of light. Under certain conditions, changes in the concentration of dissolved oxygen may affect the spectral reflectance of the water body, thereby affecting the judgment of the hyperspectral imaging system on the state of the water body.

[0083] 6. Water depth:

[0084] The depth of the water body is related to the propagation and absorption of light. As the water depth increases, the absorption and scattering effects of light by the water body increase, especially in the short-wave infrared (SWIR) and visible light bands. Through the analysis of hyperspectral data, the depth information of the water body can be inferred, especially in shallow water areas, the spectral data can be estimated by a specific algorithm to estimate the change of water depth.

[0085] 7. Substrate type (e.g. sand, mud, rock, etc.):

[0086] The water bottom sediments (such as sand, mud, rock, etc.) have different spectral characteristics, especially in the short-wave infrared (SWIR) band. The hyperspectral imaging system can help identify different types of substrates by analyzing the spectral information reflected by the water bottom, such as sand bottom, mud bottom, stone bottom, etc., which is very important for the monitoring and management of wetland water bottom.

[0087] 8. Water salinity:

[0088] The salinity of water is usually related to the change of reflectance in different bands. In the near-infrared band, the reflectance difference between saltwater and freshwater is obvious. Hyperspectral data can help distinguish between high-salinity water and freshwater, and estimate the salinity level of the water body through related spectral characteristics.

[0089] 9. Vegetation coverage and wetland vegetation health:

[0090] The health of wetland vegetation is closely related to hydrological conditions, especially water level, soil moisture, nutrient level, etc. By analyzing the reflectance of red, near-infrared and red edge bands in hyperspectral data, the health of wetland vegetation can be evaluated. The chlorophyll content, species distribution, species diversity of wetland plants can be more accurately reflected through these bands.

[0091] 10. Hydrological cycle process (evaporation and transpiration):

[0092] The hyperspectral imaging system can analyze the evaporation and transpiration process in the hydrological cycle by analyzing the spectral characteristics of the water surface reflection. Especially in the short-wave infrared (SWIR) band, the influence of water evaporation is significant, and the reflectance change can provide certain clues for the water cycle process.

[0093] Step S200: Receive the unmanned aerial vehicle timing feedback containing the wetland image with position label and time label, identify the wetland image, and extract wetland data;

[0094] The unmanned aerial vehicle acquires the wetland image in real time when working along the motion path, but does not upload the wetland image to the total control end in real time, but uploads the wetland image in a time manner, because the transmission process of the wetland image needs to consume a large amount of resources, and affects the endurance of the unmanned aerial vehicle; steps S100 to S400 are actually applied to the total control end, receive the wetland image with the position label and the time label fed back by the unmanned aerial vehicle in a time manner, identify the wetland image, and extract the wetland data, that is, the hydrological parameter in the above content.

[0095] Step S300: identifying the wetland data, determining the motion speed of the unmanned aerial vehicle, generating a motion instruction according to the motion speed, and sending the motion instruction to the unmanned aerial vehicle.

[0096] In the process of identifying the wetland data at the total control end, the motion speed of the unmanned aerial vehicle is also determined, the motion speed is converted into a motion instruction, and the motion instruction is sent to the unmanned aerial vehicle; the motion instruction is used to control the unmanned aerial vehicle to move at the motion speed.

[0097] In the technical scheme of the application, at the unmanned aerial vehicle end, the unmanned aerial vehicle identifies the collected wetland image in real time locally during the motion, adjusts the received motion instruction according to the local identification result, that is, the unmanned aerial vehicle updates the motion speed according to the local identification result; the logic of the updating process is that when an abnormal phenomenon is identified locally, the motion speed is reduced; it should be noted that the local identification process is real-time, and the motion speed is updated in real time, but every time a period of time (time) elapses, the total control end determines a motion speed according to the overall situation of the wetland area and broadcasts it, which also means that no matter how the motion speed is reduced, it will be "reset" every time a period of time elapses.

[0098] Step S400: acquiring the actual speed of the unmanned aerial vehicle according to the positioning device built in the unmanned aerial vehicle, comparing the actual speed with the motion speed, marking the acquisition time, reading the image of the marked acquisition time when the unmanned aerial vehicle returns.

[0099] In the technical scheme of the application, the unmanned aerial vehicle does not need to upload data, it only does not need to upload the image (large amount of data) in real time, and the position needs to be uploaded, the uploading process does not need to consume too many resources, and the position can be obtained according to the positioning device built in the unmanned aerial vehicle, after the position is obtained, the actual speed can be calculated, the actual speed is compared with the motion speed, the position where the actual speed is slow indicates that the unmanned aerial vehicle detects an abnormality at the corresponding time, at this time, the corresponding time is marked, and the process is continuously executed, the position can be used to remotely determine the time when the abnormality exists, when the unmanned aerial vehicle returns, the local database of the unmanned aerial vehicle is accessed, and the image of the marked acquisition time is read.

[0100] Figure 2The first sub-process block diagram of the ecological wetland data monitoring method based on the hyperspectral imaging system, the map of the wetland area is obtained, and the motion path of the unmanned aerial vehicle is determined according to the map, and the steps of synchronously determining the motion path of the unmanned aerial vehicle comprise:

[0101] Step S101: a map containing a wetland area is obtained and displayed, a user input region boundary is received, and a wetland area is determined;

[0102] Step S102: receiving user input of the working parameters of the unmanned aerial vehicle, determining the cutting width according to the working parameters; the working parameters include the working height and the working wide angle;

[0103] Step S103: uniformly setting sampling points on the region boundary, sequentially querying the farthest sampling point of each sampling point, and obtaining the distance thereof;

[0104] Step S104: selecting and connecting the two sampling points with the largest distance as the direction line;

[0105] Step S105: based on the cutting width array, the direction line is calculated, the intersection of the direction line after the array and the wetland area is calculated, the intersection is sequentially connected from the beginning to the end, and the motion path is obtained.

[0106] In an example of the technical scheme of the present application, a map containing a wetland area is obtained and displayed, a user input region boundary is received, and a wetland area is determined; user input of the working parameters of the unmanned aerial vehicle is received, and the cutting width is determined according to the working parameters. In actual application, the unmanned aerial vehicle generally collects the position directly below, and the working parameters include the working height and the working wide angle. The collected area is actually a circle, and the radius of the circle is the cutting width. Sampling points are uniformly set on the region boundary, the farthest sampling point of each sampling point is sequentially queried, and the distance thereof is obtained. By comparing the calculated distance, the two farthest sampling points can be calculated. The two farthest sampling points are connected as a reference line segment, also known as a direction line. Based on the cutting width array, the direction line after the array and the wetland area are calculated, the intersection is sequentially connected from the beginning to the end, and the motion path is obtained.

[0107] It is worth mentioning that the cutting width in the above content can also be simply corrected, that is, 60% (or other values) of the radius of the circle is selected as the cutting width, which means that the collection areas on the adjacent direction lines after the array have a part of overlap. At this time, by comparing the overlapping part, it can be judged whether there is a problem in the detection process of the unmanned aerial vehicle.

[0108] Figure 3 The second sub-process block diagram of the ecological wetland data monitoring method based on the hyperspectral imaging system, the wetland image containing position labels and time labels received by the unmanned aerial vehicle is identified, and the wetland data is extracted, and the steps comprise:

[0109] Step S201: send the motion path to at least one unmanned aerial vehicle, and receive the wetland images fed back by the unmanned aerial vehicle, the wetland images containing position tags and time tags;

[0110] Step S202: determine the wetland images of different wave bands at the latest time of each position according to the position tags and the time tags;

[0111] Step S203: splice all the wetland images to obtain the latest images of the entire wetland area for each wave band;

[0112] Step S204: identify the latest images of each wave band, record the identification time, and obtain the wetland data containing time tags; the wetland data is an array, and each element corresponds to a parameter.

[0113] In one example of the technical scheme, the motion path is sent to at least one unmanned aerial vehicle, and the wetland images fed back by the unmanned aerial vehicle are received, the wetland images containing position tags and time tags, indicating the wetland images obtained at which position at which time; then, the wetland images at the latest time at each position are inquired, and splicing is performed to obtain the latest images of the entire wetland area; for each wave band, such an operation needs to be performed once, and therefore, multiple latest images can be obtained.

[0114] The latest images are identified, the identification time is recorded, and the wetland data containing time tags is obtained; in the identification process, an existing identification scheme can be adopted, and the target of identification is the hydrological parameters mentioned above.

[0115] Figure 4 The third sub-flow block diagram of the ecological wetland data monitoring method based on the hyperspectral imaging system, the step of identifying the wetland data, determining the motion speed of the unmanned aerial vehicle, generating a motion instruction according to the motion speed, and sending the motion instruction to the unmanned aerial vehicle comprises:

[0116] Step S301: arrange the wetland data containing time tags in a time sequence to obtain a data matrix;

[0117] Step S302: cut the data matrix according to a preset time span, compare the cut data matrix with a preset standard data set, and determine the motion speed of the unmanned aerial vehicle according to a comparison result;

[0118] Step S303: generate a motion instruction according to the motion speed, and send the motion instruction to the unmanned aerial vehicle;

[0119] In an example of the technical solution of the present application, the wetland data containing time tags is arranged in time sequence to obtain a data matrix. The reason why it is a data matrix is that the wetland data is not unique and includes many (hydrological parameters in the above content). Therefore, the wetland data is actually an array, which is expanded into a two-dimensional array according to time, that is, a data matrix. The rows in the data matrix represent time, and the columns represent a certain parameter. The data matrix is intercepted according to a preset time span, the intercepted data matrix is compared with a preset standard data set, the motion speed of the unmanned aerial vehicle is determined according to the comparison result, the standard data set is a standard state set in advance, the difference between the intercepted data matrix and the standard data set is calculated, and the motion speed is determined according to the difference. After the motion speed is determined, the motion instruction is generated according to the motion speed, and the unmanned aerial vehicle is sent.

[0120] The determination process of the motion speed is as follows:

[0121] In the formula, The motion speed is represented by V, and respectively represent the intercepted data matrix and the standard data set, represents and the number of intersection elements; the determination process of the intersection elements is as follows: when the difference between the data in the intercepted data matrix and the corresponding data in the standard data set is less than a preset threshold value, the data is taken as an intersection element; The total number of elements of the intercepted data matrix is represented by N; is a preset correction coefficient.

[0122] The calculation principle of the above content is not complex. For the intercepted data matrix, it is judged whether each data in the intercepted data matrix is in a standard state based on the data in the standard data set. Whether the current wetland data is in a standard state is determined according to the number of data in the standard state. The more the number is, the more standard it is, and the faster the motion speed of the unmanned aerial vehicle is, that is, the faster the unmanned aerial vehicle is.

[0123] Figure 5 The fourth sub-process block diagram of the ecological wetland data monitoring method based on the hyperspectral imaging system is as follows: the actual speed of the unmanned aerial vehicle is obtained according to the positioning device built in the unmanned aerial vehicle, the actual speed and the motion speed are compared, the acquisition time is marked, and when the unmanned aerial vehicle returns, the image of the marked acquisition time is read.

[0124] Step S401: The position of the unmanned aerial vehicle is obtained in real time according to the positioning device built in the unmanned aerial vehicle.

[0125] Step S402: The actual speed of the unmanned aerial vehicle is calculated according to the position.

[0126] Step S403: when the actual speed is less than the motion speed, mark the corresponding time;

[0127] Step S404: when the UAV returns, read the image of the marked collection time.

[0128] In an example of the technical scheme of the present application, the marking process of the time is limited, the position of the UAV is acquired in real time according to the positioning device built in the UAV, the actual speed of the UAV is calculated according to the position, the actual speed is compared with the motion speed of the current time, when the actual speed is less than the motion speed, the corresponding time is marked, finally, when the UAV returns, the image of the marked collection time is read.

[0129] It should be noted that the motion speed of the current time refers to the motion speed of the nearest time sent by the total control end to the UAV.

[0130] In an example of the technical scheme of the present application, the working process of the UAV includes:

[0131] receiving a motion instruction;

[0132] real-time collection of wetland images, positioning of abnormal features and their levels in the wetland images based on a preset convolution recognition model;

[0133] determining the risk degree of the wetland images according to the levels and quantities of the abnormal features;

[0134] adjusting the motion instruction according to the risk degree; wherein the greater the risk degree, the smaller the speed corresponding to the adjusted motion instruction.

[0135] In an example of the technical scheme of the present application, the working process of the UAV is specifically limited, which includes: receiving a motion instruction, real-time collection of wetland images, positioning of abnormal features and their levels in the wetland images based on a preset convolution recognition model, the convolution recognition model is essentially an image comparison model, some image features of abnormal objects or conditions are pre-counted, in actual application, the image features are selected in turn to traverse the wetland images, and matching is continuously performed, so that the abnormal features in the wetland images can be positioned, in the technical scheme of the present application, each abnormal feature is also pre-labeled with a level to represent the abnormal degree of each abnormal object or condition; finally, the risk degree of the wetland images is determined according to the levels and quantities of the abnormal features, and the motion instruction is adjusted according to the risk degree.

[0136] Figure 6 The composition structure block diagram of the ecological wetland data monitoring system based on the hyperspectral imaging system, in the embodiment of the present application, an ecological wetland data monitoring system based on a hyperspectral imaging system, the system 10 comprises:

[0137] The motion path determination module 11 is configured to acquire a map of the wetland area, select a UAV according to the map, and determine a motion path of the UAV synchronously; and the UAV is provided with a hyperspectral imaging system.

[0138] The wetland data extraction module 12 is configured to receive wetland images with position labels and time labels fed back by the UAV in a timely manner, identify the wetland images, and extract wetland data.

[0139] The wetland data identification module 13 is configured to identify the wetland data, determine a motion speed of the UAV, generate a motion instruction according to the motion speed, and send the motion instruction to the UAV.

[0140] The marker reading module 14 is configured to acquire an actual speed of the UAV according to a positioning device built in the UAV, compare the actual speed with the motion speed, and read an image of a collection time of the marker when the UAV returns.

[0141] In the process of motion, the UAV performs real-time local identification on the collected wetland images, and adjusts the received motion instruction according to the local identification result.

[0142] Further, the motion path determination module 11 comprises:

[0143] The area determination unit is configured to acquire and display a map containing a wetland area, receive a region boundary input by a user, and determine the wetland area.

[0144] The width determination unit is configured to receive a working parameter of the UAV input by the user, and determine a segmentation width according to the working parameter; the working parameter comprises a working height and a working wide angle.

[0145] The distance acquisition unit is configured to set sampling points on the region boundary evenly, query a farthest sampling point of each sampling point in sequence, and acquire a distance of the farthest sampling point.

[0146] The direction line determination unit is configured to select and connect two sampling points with the largest distance as a direction line.

[0147] The connection unit is configured to calculate an intersection of the direction line and the wetland area after arraying the direction line based on the segmentation width array, and connect the intersection in sequence to obtain the motion path.

[0148] Specifically, the wetland data extraction module 12 comprises:

[0149] The image receiving unit is configured to send the motion path to at least one UAV, and receive wetland images with position labels and time labels fed back by the UAV.

[0150] An image selecting unit is configured to determine the latest time of each location of the different wave bands of the wetland images according to the location label and the time label;

[0151] An image splicing unit is configured to splice all the wetland images to obtain the latest image of the whole wetland area for each wave band;

[0152] A recording unit is configured to identify the latest image of each wave band, record the identification time, and obtain the wetland data containing the time label; the wetland data is an array, and each element corresponds to a parameter.

[0153] Further, the wetland data identification module 13 comprises:

[0154] A data arrangement unit is configured to arrange the wetland data containing the time label according to the time sequence to obtain a data matrix;

[0155] A comparison execution unit is configured to intercept the data matrix according to a preset time span, compare the intercepted data matrix with a preset standard data group, and determine the motion speed of the unmanned aerial vehicle according to the comparison result;

[0156] An instruction generation and sending unit is configured to generate a motion instruction according to the motion speed and send the motion instruction to the unmanned aerial vehicle;

[0157] The determination process of the motion speed is as follows:

[0158] ; wherein, represents the motion speed, and respectively represent the intercepted data matrix and the standard data group, represents and the number of intersection elements; the determination process of the intersection elements is as follows: when the difference between the data in the intercepted data matrix and the corresponding data in the standard data group is less than a preset threshold, the data is taken as an intersection element; represents the total number of elements of the intercepted data matrix; is a preset correction coefficient.

[0159] The above merely describes the preferred embodiments of the present application and is not used to limit the present application, and any modification, equivalent replacement, and improvement made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for monitoring ecological wetland data based on a hyperspectral imaging system, characterized in that, The method comprises: acquiring a map of a wet area, selecting a UAV according to the map, and synchronously determining a movement path of the UAV; the UAV is provided with a hyperspectral imaging system; receiving wet images with position tags and time tags fed back by the UAV in a time sequence, identifying the wet images, and extracting wet data; identifying the wet data, determining a movement speed of the UAV, generating a movement instruction according to the movement speed, and sending the movement instruction to the UAV; acquiring an actual speed of the UAV according to a positioning device built in the UAV, comparing the actual speed with the movement speed to mark a collection time, and reading images of the marked collection time when the UAV returns; wherein the UAV performs real-time local identification on the collected wet images during movement, and adjusts the received movement instruction according to the local identification result; the step of identifying the wet data, determining the movement speed of the UAV, generating the movement instruction according to the movement speed, and sending the movement instruction to the UAV comprises: arranging the wet data with time tags in a time sequence to obtain a data matrix; cutting the data matrix according to a preset time span, comparing the cut data matrix with a preset standard data set, and determining the movement speed of the UAV according to a comparison result; generating the movement instruction according to the movement speed, and sending the movement instruction to the UAV; wherein the determination process of the movement speed comprises: wherein, represents the moving speed, and respectively represent the intercepted data matrix and the standard data set, represents and the number of intersection elements of the two sets; the determination process of the intersection elements is that when the difference between the data in the intercepted data matrix and the corresponding data in the standard data set is less than a preset threshold, the data is regarded as an intersection element; represents the total number of elements in the intercepted data matrix; is a preset correction coefficient.

2. The hyperspectral imaging system based ecological wetland data monitoring method according to claim 1, characterized in that, the step of acquiring the map of the wet area, selecting the UAV according to the map, and synchronously determining the movement path of the UAV comprises: acquiring and displaying a map containing the wet area, receiving a region boundary input by a user, and determining the wet area; receiving working parameters of the UAV input by the user, determining a segmentation width according to the working parameters; the working parameters include a working height and a working wide angle; uniformly setting sampling points on the region boundary, sequentially querying a farthest sampling point of each sampling point, and acquiring a distance thereof; selecting and connecting two sampling points with the largest distance as a direction line; based on the direction line and the segmentation width array, calculating an intersection of the direction line after arraying and the wet area, sequentially connecting the intersection from a head to a tail to obtain the movement path. 3.The method of monitoring ecological wetland data based on a hyperspectral imaging system according to claim 1, wherein, the step of receiving the wet images with position tags and time tags fed back by the UAV in a time sequence, identifying the wet images, and extracting the wet data comprises: sending the movement path to at least one UAV, and receiving wet images with position tags and time tags fed back by the UAV; determining wet images of different bands at a latest time of each position according to the position tags and the time tags; for each band, splicing all the wet images to obtain a latest image of the entire wet area; identifying the latest image of each band, recording an identification time, and obtaining wet data with time tags; the wet data is an array, and each element corresponds to one parameter.

4. The hyperspectral imaging system based ecological wetland data monitoring method according to claim 1, wherein, the step of acquiring the actual speed of the UAV according to the positioning device built in the UAV, comparing the actual speed with the movement speed to mark the collection time, and reading the images of the marked collection time when the UAV returns comprises: acquiring a position of the UAV in real time according to the positioning device built in the UAV; calculating the actual speed of the UAV according to the position. When the actual speed is less than the motion speed, mark the corresponding moment; When the unmanned aerial vehicle returns, read the image of the collection moment of the mark.

5. The hyperspectral imaging system based ecological wetland data monitoring method according to claim 1, wherein, The working process of the unmanned aerial vehicle includes: receiving a motion instruction; collecting a wetland image in real time, positioning an abnormal feature and a level thereof in the wetland image based on a preset convolution recognition model; determining a risk degree of the wetland image according to the level and the number of the abnormal feature; adjusting the motion instruction according to the risk degree; the greater the risk degree, the smaller the speed corresponding to the adjusted motion instruction.

6. An ecological wetland data monitoring system based on a hyperspectral imaging system, characterized in that, The system includes: a motion path determination module, configured to obtain a map of a wetland area, select an unmanned aerial vehicle according to the map, and determine a motion path of the unmanned aerial vehicle synchronously; the unmanned aerial vehicle is internally provided with a hyperspectral imaging system; a wetland data extraction module, configured to receive wetland images with position labels and time labels fed back by the unmanned aerial vehicle at a regular time, identify the wetland images, and extract wetland data; a wetland data identification module, configured to identify the wetland data, determine a motion speed of the unmanned aerial vehicle, generate a motion instruction according to the motion speed, and send the motion instruction to the unmanned aerial vehicle; a mark reading module, configured to obtain an actual speed of the unmanned aerial vehicle according to a positioning device internally provided in the unmanned aerial vehicle, compare the actual speed with the motion speed, mark a collection moment, and read an image of the collection moment of the mark when the unmanned aerial vehicle returns; wherein the unmanned aerial vehicle performs real-time local identification on the collected wetland image during the motion, and adjusts the received motion instruction according to the local identification result; the wetland data identification module includes: a data arrangement unit, configured to arrange the wetland data with the time labels according to a time sequence to obtain a data matrix; a comparison execution unit, configured to cut the data matrix according to a preset time span, compare the cut data matrix with a preset standard data group, and determine the motion speed of the unmanned aerial vehicle according to a comparison result; an instruction generation and sending unit, configured to generate the motion instruction according to the motion speed, and send the motion instruction to the unmanned aerial vehicle; wherein the determination process of the motion speed is as follows: wherein, represents the moving speed, and respectively represent the intercepted data matrix and the standard data set, represents and the number of intersection elements; the determination process of the intersection elements is that when the difference between the data in the intercepted data matrix and the corresponding data in the standard data set is less than a preset threshold, the data is taken as an intersection element; represents the total number of elements in the intercepted data matrix; is a preset correction coefficient.

7. The hyperspectral imaging system based ecological wetland data monitoring system according to claim 6, wherein, the motion path determination module includes: a region determination unit, configured to obtain and display a map containing a wetland area, receive a region boundary input by a user, and determine the wetland area; a width determination unit, configured to receive working parameters of the unmanned aerial vehicle input by the user, and determine a segmentation width according to the working parameters; the working parameters include a working height and a working wide angle; a distance acquisition unit, configured to evenly set sampling points on the region boundary, sequentially query the farthest sampling point of each sampling point, and acquire a distance thereof; a direction line determination unit, configured to select and connect two sampling points with the largest distance as a direction line; a connection unit, configured to calculate an intersection of the direction line after arraying and the wetland area based on the arrayed direction line, sequentially connect the intersection from the head to the tail, and obtain the motion path.

8. The hyperspectral imaging system based ecological wetland data monitoring system according to claim 6, wherein, the wetland data extraction module includes: an image receiving unit, configured to send the motion path to at least one unmanned aerial vehicle, and receive wetland images with position labels and time labels fed back by the unmanned aerial vehicle; an image selection unit, configured to determine different-band wetland images of the latest moment of each position according to the position labels and the time labels; and an image processing unit, configured to perform image processing on the wetland images. An image splicing unit is configured to splice all the wetland images for each wave band to obtain a latest image of the whole wetland area. A recording unit is configured to identify the latest image of each wave band, record the identification time, and obtain wetland data containing a time label; the wetland data is an array, and each element corresponds to a parameter.

Images