A route planning method based on multi-band camouflage feature data collection

Abstract

The present invention discloses a route planning method based on multi-band camouflage feature data collection, and belongs to the technical field of unmanned aerial vehicle (UAV) route planning. First, route planning is performed according to the type of UAV and the camera carried. A multi-rotor UAV equipped with a visible light or infrared camera considers the impact of different weather conditions on the resolution of photos collected by the UAV, adjusts the flight altitude of the UAV, and flies around all targets to obtain the optimal flying order. A fixed-wing UAV equipped with a synthetic aperture radar performs multi-perspective route planning and full coverage route planning on the target area. The routes of the above two types of UAVs are respectively converted into coordinates to obtain the latitude and longitude coordinates of the UAV waypoints, and finally obtain the route file to plan the flight route of the UAV. The present invention takes into account a variety of weather factors, so that the UAV can collect camouflage feature data of better quality in complex weather environments.

Description

Technical Field

[0001] The invention belongs to the technical field of unmanned aerial vehicle route planning, and specifically refers to a route planning method based on multi-band camouflage feature data collection. Background Art

[0002] As drone technology becomes increasingly mature, drones have played a huge role in surveying, combat, search and rescue, and data collection. In complex environments and scenarios, it is critical to achieve autonomous planning and navigation of drone routes. In real environments, weather factors such as fog, temperature, humidity, and carbon dioxide concentration can affect the quality of data collected by drones. Existing drone route planning methods do not take into account the impact of weather on routes, nor do they plan routes based on the different bands of data collection equipment carried by drones.

[0003] Common UAV route planning methods include A_star, D_star, RRT, ant colony algorithm, genetic algorithm, particle swarm algorithm and artificial potential field method. When applying these algorithms or improving them to solve route planning problems, the impact of weather factors on route planning is not taken into account, nor is route planning for specific tasks for UAVs equipped with multi-band data acquisition equipment such as visible light, infrared or SAR (synthetic aperture radar). Summary of the invention

[0004] The technical problem to be solved by the present invention is to realize autonomous route planning of UAVs in different bands and optimize the routes of UAVs under the influence of weather factors, so that the UAVs can collect camouflage feature data with better quality.

[0005] The route planning method based on multi-band camouflage feature data collection has the following specific steps:

[0006] Step 1: According to the type of drone and the data collection equipment it carries, the multi-rotor drone equipped with a visible light or infrared camera among the drones for route planning shall execute step 2; the fixed-wing drone equipped with a synthetic aperture radar shall execute step 4;

[0007] Step 2: Consider the impact of weather on multi-rotor drones equipped with visible light or infrared cameras, calculate the actual flight altitude of the drones for data collection on the targets, and plan the optimal order for the drones to fly around the targets;

[0008] (1) When the UAV collects photos vertically from the ground, the theoretical flight altitude H of the UAV is calculated by the parameters of the camera payload, and the flight altitude is adjusted according to the weather conditions to obtain the actual flight altitude h of the UAV;

[0009] The theoretical flight altitude H of the drone is:

[0010]

[0011] The parameters of the camera payload include the focal length of the camera lens f, the pixel size of the camera a, and the ground resolution GSD of the photo.

[0012] Due to the influence of weather, the resolution of the onboard camera for shooting the target is guaranteed by changing the flight altitude of the drone, specifically:

[0013] ①Fog will weaken the reflected light of the target, and the light will be scattered into the camera by the fog. Then, under foggy weather conditions, the light intensity t(x) received by the drone's onboard camera is:

[0014] t(x)=e -γd(x)

[0015]

[0016] Among them, d(x) is the depth of field, that is, the height of the lens from the ground when the drone collects data; γ is the scattering coefficient caused by fog; λ represents the wavelength (in microns), and Q represents the visibility of the fog (in meters).

[0017] The light intensity t(x) reaching the airborne camera decays exponentially with the increase of the fog scattering coefficient γ. The flight altitude h of the drone is equivalent to the depth of field d(x), so the actual flight altitude of the drone is:

[0018] h=ln(t(x)) / γ

[0019] If the light intensity t(x) received by the camera is to be kept constant, the flying altitude of the UAV is lowered to ensure the clarity of the captured image and the distinguishability of the target.

[0020] ②The influence of humidity and carbon dioxide on the atmospheric transmittance of infrared cameras is:

[0021] Infrared atmospheric transmittance = radiation flux after propagation attenuation / radiation flux before incident

[0022] When the infrared wavelength is 8-12μm, the influence of water vapor and carbon dioxide on the atmospheric transmittance τ of infrared radiation is exponential, specifically:

[0023]

[0024]

[0025] When the atmospheric humidity is greater than 85%, the drone equipped with an infrared camera stops shooting the target; when the atmospheric humidity is less than 85%, the flight altitude of the drone is adjusted manually according to the actual situation. The higher the humidity, the lower the actual flight altitude.

[0026] ③At the same time, the atmospheric transmittance and ambient temperature affect the target measurement temperature, further affecting the data collection effect of the infrared camera on the target;

[0027] The measured temperature of the target t c for:

[0028]

[0029] Where τ represents the atmospheric transmittance, represents the target surface emissivity, t0 represents the actual temperature of the target being measured, t a Indicates the ambient temperature.

[0030] The greater the temperature difference between the ambient temperature and the actual temperature of the target, and the greater the target measured temperature, the best data collection effect on the target.

[0031] (2) The UAV obtains the circumnavigation radius of each target based on the actual flight altitude;

[0032] Assume that the length and width of the target are L and W respectively, the distance between the camera lens and the target is S, the actual flight height of the drone perpendicular to the ground is h, the flight radius is r, the proportion of the target in the target photo taken by the camera is y, the resolution of the photo is p (cm / pixel) and the photo size is m×n, then the radius r of the drone flying around the target is:

[0033]

[0034] The distance S between the camera lens and the target is:

[0035]

[0036] The photo ground resolution GSD is calculated based on the photo resolution, photo size, and target length and width:

[0037]

[0038] (3) The specific process of determining the optimal order of flying around targets is as follows:

[0039] Step 201, input the longitude and latitude of the take-off point of the drone and each target into the drone route planning program. The drone takes off from the take-off point s, and the final destination is i. According to the calculated actual flight altitude and circling radius, the drone circling each target, and calculates the shortest path length d(i,v) from the destination i through each target in V (a set of other target points except the target point i) once and only once, and finally returns to the take-off point s;

[0040] The distance from the destination point i to any destination point k in the set V is expressed as c ik express;

[0041] Step 202, determine whether the set V is an empty set. If so, d(i, v) means directly returning from the destination point i to the starting point s. At this time, d(i, v) = c is And i≠s; when the set V is not an empty set, the UAV attempts to fly around each target in the set V and finds the shortest flying distance, which is the optimal order of flying around the targets;

[0042] The optimal order for a drone to fly around multiple targets is to solve the equation:

[0043]

[0044] Step 203, the UAV collects data on all targets at different altitudes according to the optimal circling sequence for each target;

[0045] The process of the drone collecting data on various targets at different altitudes is as follows: for a certain target, the drone flies around at a rotation angle interval of α, and collects a photo of the target every time the drone rotates by an angle α until it circles the entire target and collects photos from multiple angles.

[0046] Step 3: Set the latitude and longitude coordinates of each target [lng i ,lat i ] is converted to plane coordinates [x i ,y i ], and calculate the plane coordinates of each waypoint corresponding to each target according to the flying radius and rotation angle of the UAV, and enter step five;

[0047] Plane coordinates of each waypoint corresponding to each target It is expressed as:

[0048]

[0049]

[0050] Among them, j represents the j-th waypoint number.

[0051] Step 4: The fixed-wing UAV equipped with synthetic aperture radar scans each target area of ​​the route planning to obtain the plane coordinates of the scanned area;

[0052] Specifically:

[0053] Step 401: input the coordinates of the center point of the scanning area corresponding to the current target, the width L1 and length L2 of the rectangular scanning area, the UAV flight altitude H', the ground grazing angle β, and the width W' of the UAV strip coverage area into the synthetic aperture radar;

[0054] If the drone performs multi-view route planning, the scanning area is a square, that is, L1 = L2;

[0055] Step 402: Based on the width L1 and length L2 of the rectangular scanning area, the flight height H', the ground rubbing angle β and the coordinates (X O ,Y O ), calculate the vertex coordinates of the rectangular scanning area, and obtain the quadrilateral flight route of the UAV.

[0056] If it is a multi-view route planning, a rectangle is made with the diagonal length 7 / 5L1 of the square scanning area, with the east direction as the positive direction of the X axis. The coordinates of the four vertices of the rectangle are:

[0057]

[0058]

[0059]

[0060]

[0061] If the route is planned for full coverage, with the east direction as the positive direction of the X axis, the coordinates of the four vertices of the rectangular scanning area are:

[0062]

[0063] X n =X1+(u-1)×F

[0064]

[0065]

[0066] Among them, F is the route spacing, and 500 is the length reserved for the start and end points of each route. u represents the uth route. The fixed-wing UAV equipped with synthetic aperture radar flies back and forth, and eventually there will be U routes in a rectangular area.

[0067] Step 5, converting the plane coordinates of each waypoint obtained in step 3 or step 4 into longitude and latitude coordinates, obtaining each route waypoint of the drone flying around the target, and generating a specific route file;

[0068] Step 6: Import the route file into the flight control system of the drone, plan the actual flight route of the drone, control the drone to fly along the set route and collect data.

[0069] The beneficial effects of the present invention are:

[0070] The present invention takes into account for the first time the constraints of various weather factors on UAV route planning, and autonomously plans the route of a UAV equipped with one or two data acquisition devices including visible light, infrared cameras or synthetic aperture radars, so that the UAV can collect camouflage feature data of better quality in complex weather environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 It is a schematic diagram of the flying radius of the UAV of the present invention around the target;

[0072] Figure 2 is a flow chart of coordinate conversion of the drone of the present invention;

[0073] Figure 3 Schematic diagram of optional circumnavigation paths between the drone of the present invention and multiple target points;

[0074] Figure 4 This is a schematic diagram of selecting the shortest detour route for the UAV of the present invention;

[0075] Figure 5 This is a schematic diagram of multi-view route planning for a fixed-wing UAV equipped with a synthetic aperture radar according to the present invention;

[0076] Figure 6 It is a schematic diagram of the scanning area for multi-view route planning of a fixed-wing UAV equipped with a synthetic aperture radar according to the present invention;

[0077] Figure 7 It is a schematic diagram of the full coverage route planning principle of a fixed-wing UAV equipped with a synthetic aperture radar according to the present invention;

[0078] Figure 8 It is a schematic diagram of the scanning area for full coverage route planning of a fixed-wing UAV equipped with a synthetic aperture radar according to the present invention. DETAILED DESCRIPTION

[0079] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0080] The present invention considers the flight path planning of multi-rotor UAVs and the SAR (synthetic aperture radar) route planning of fixed-wing UAVs. For the flight path planning of multi-rotor UAVs, two different loads of visible light and infrared cameras are considered, and the flight targets are considered to be a single target and multiple targets. The SAR route planning of fixed-wing UAVs considers multi-view route planning of a single target and full coverage route planning.

[0081] A route planning method based on multi-band camouflage feature data collection, the specific steps are as follows:

[0082] Step 1: According to the type of drone and the data collection equipment it carries, the multi-rotor drone equipped with a visible light or infrared camera among the drones for route planning shall execute step 2; the fixed-wing drone equipped with a synthetic aperture radar shall execute step 4;

[0083] Step 2: Consider the impact of weather on multi-rotor drones equipped with visible light or infrared cameras, calculate the actual flight altitude of the drones for data collection on the targets, and plan the optimal order for the drones to fly around the targets;

[0084] (1) When the UAV collects photos vertically from the ground, the theoretical flight altitude H of the UAV is calculated by the parameters of the camera payload, and the flight altitude is adjusted according to the weather conditions to obtain the actual flight altitude h of the UAV;

[0085] The theoretical flight altitude H of the drone is:

[0086]

[0087] The parameters of the camera payload include the focal length of the camera lens f, the pixel size of the camera a, and the ground resolution GSD of the photo.

[0088] Due to the influence of different weather conditions, when the UAV is at the theoretical flight altitude, the resolution of the onboard camera to the target is reduced and cannot meet the data collection requirements. Therefore, under different weather conditions, the actual flight altitude of the UAV is changed to ensure the resolution of the onboard camera to the target. Specifically:

[0089] 1) The impact of foggy weather on the actual flight altitude of multi-rotor drones equipped with visible light cameras;

[0090] When building the model, the fog level is used as an input condition for route planning. In the weather forecast system, horizontal visibility below 1 km is called fog, horizontal visibility between 200-500 meters is called heavy fog, horizontal visibility between 50-200 meters is called dense fog, and horizontal visibility less than 50 meters is called severe fog. When in dense fog or severe fog, it is not suitable for drone data collection because of the influence of water vapor on the camera lens and the lack of clarity of the photos taken for post-processing; when in other types of foggy days, the visibility of the fog is used as a quantitative input condition, and the visibility of the fog is roughly measured with the help of the weather forecast system or the fog visibility measurement sensor.

[0091] The atmospheric light model is the basis of the current defogging algorithm. The model believes that the formation of a foggy image is composed of the light reflected by the object weakened by the fog and the atmospheric light reflected by the fog entering the camera. In the present invention, it is set that the weakening and scattering of the fog are related to the depth of field from the object to the camera, and are inversely proportional to the light intensity reaching the camera.

[0092] In foggy weather conditions, the light intensity t(x) received by the drone's onboard camera is:

[0093] t(x)=e -γd(x)

[0094]

[0095] Among them, d(x) is the depth of field, that is, the height of the lens from the ground when the drone collects data; γ is the scattering coefficient caused by fog; λ represents the wavelength (in microns), and Q represents the visibility of fog (in meters). In foggy conditions, the scattered light on the surface of the object is lost due to the scattering of fog, the resolution and color of the photo change, the light intensity reaching the camera decreases and decays exponentially with the increase of the propagation distance.

[0096] According to the above theoretical analysis, the light intensity reaching the camera decays exponentially. The flight altitude h of the UAV is equivalent to the depth of field d(x). Then the flight altitude of the UAV is h = ln(t(x)) / γ. While keeping t(x) unchanged, the flight altitude of the UAV can be reduced to ensure the clarity of the captured image and the discernibility of the target.

[0097] For different fog visibility, the scattering coefficient γ is different. Through actual experiments, the corresponding values ​​of fog visibility and scattering coefficient can be measured. Therefore, in actual data collection, the flight altitude of the UAV can be lowered while ensuring the safety of the UAV, so that clearer optical image data can be collected.

[0098] 2) The impact of humidity and carbon dioxide on the atmospheric transmittance of the infrared camera carried by the UAV;

[0099] When electromagnetic waves propagate in the atmosphere, atmospheric transmittance is an important factor affecting infrared radiation transmission. Atmospheric transmittance is the ratio of electromagnetic radiation flux after atmospheric attenuation to electromagnetic radiation flux at the time of incidence. Infrared is also an electromagnetic wave. When propagating in the atmosphere, it will also be affected by various substances in the atmosphere. Therefore, the atmospheric transmittance of infrared is:

[0100] Infrared atmospheric transmittance = radiation flux after propagation attenuation / radiation flux before incident

[0101] The factors affecting the atmospheric transmittance are mainly water vapor, carbon dioxide and ozone. The present invention mainly considers the influence of water vapor (humidity) and carbon dioxide.

[0102] The following is an empirical formula for the effect of water vapor and carbon dioxide on the atmospheric transmittance of light of different wavelengths, which represents the attenuation of light waves by water vapor and carbon dioxide. It can be seen that when the infrared wavelength is 8-12μm (the commonly used band for infrared cameras), the atmospheric transmittance decays exponentially:

[0103]

[0104]

[0105] Humidity is a condition that affects infrared photography. When the atmospheric humidity is greater than 85%, the attenuation of the atmosphere's radiation to the target object increases sharply due to the increase in water vapor density and the increase in water vapor absorption of infrared radiation, making it unsuitable for infrared data collection. When the atmospheric humidity is less than 85%, the flight altitude is optimized according to the empirical formula in 2).

[0106] According to the theoretical model established above, when the UAV collects infrared data in an environment with high humidity and carbon dioxide concentration, the quality of the collected data can be improved by lowering the flight altitude of the UAV while ensuring the safety of the UAV.

[0107] 3) At the same time, the atmospheric transmittance and ambient temperature affect the target measurement temperature, further affecting the data collection effect of the infrared camera on the target;

[0108] Temperature is an input condition. Temperature is mainly reflected in its impact on infrared photography. Before being received, infrared radiation must pass through media such as the atmosphere and imaging system, resulting in infrared loss. According to the Steve-Boltzmann law, the total emissivity of a black body is proportional to the fourth power of the black body's thermodynamic temperature. Therefore, the higher the temperature, the stronger the infrared rays emitted by the object. Therefore, within a certain range, the higher the object temperature, the more conducive it is to infrared detection, and the lower the surface temperature, the greater the temperature difference between the object and the surface, and the more obvious the difference between the object and the background in the infrared photo taken. Therefore, the best data collection effect for infrared photos is in the morning or at night when the temperature difference is large.

[0109] The measured temperature of the target t c for:

[0110]

[0111] Where τ represents the atmospheric transmittance, represents the target surface emissivity, t0 represents the actual temperature of the target being measured, t a Indicates the ambient temperature.

[0112] (2) The calculation process of the UAV's flying radius r around the target is:

[0113] like Figure 1 As shown in the figure, assuming that the length and width of the target are L and W respectively, the distance between the camera lens and the target is S, the actual flight height of the drone perpendicular to the ground is h, the flight radius is r, the target proportion in the target photo obtained by the camera is y, the resolution of the photo is p (cm / pixel) and the photo size is m×n, then the flying radius r of the drone around the target is:

[0114]

[0115] The distance S between the camera lens and the target is:

[0116]

[0117] The photo ground resolution GSD is calculated based on the photo resolution, photo size, and target length and width:

[0118]

[0119] (3) The specific process of determining the optimal order for drones to fly around targets is as follows:

[0120] Step 201, input the longitude and latitude of the take-off point of the drone and each target into the drone route planning program. The drone takes off from the take-off point s, and the final destination is i. According to the calculated actual flight altitude and circling radius, the drone circling each target, and calculates the shortest path length d(i,v) from the destination i through each target in V (a set of other target points except the target point i) once and only once, and finally returns to the take-off point s;

[0121] The distance from the destination point i to any destination point k in the set V is expressed as c ik express;

[0122] Step 202, determine whether the set V is an empty set. If so, then d(i, v) means directly returning from the destination point i to the starting point s. At this time, d(i, v) = c is and i≠s; when the set V is not an empty set, in the set V, the drone tries every circling sequence and finds the shortest circling distance, which is the optimal sequence for circling the target;;

[0123] like Figure 3 As shown in the figure, "UAV" represents the take-off point of the UAV, and "1, 2, 3, 4, 5, 6" represent multiple targets. Given the take-off point of the UAV and the longitude and latitude of each flying target, the UAV traverses each target once from the take-off point and returns to the take-off point. This problem is a "traveling salesman problem". The present invention proposes a dynamic programming method to solve this problem, and abstracts the problem to obtain a dynamic programming equation, namely: the UAV starts from the take-off point s, d(i, v) represents the shortest path length from vertex i through each vertex in V (a set of target points) once and only once, and finally returns to the take-off point s of the UAV, c ik represents the distance between the currently selected target point and target i, d(k,V-{k}) is a subproblem.

[0124] The optimal sequence dynamic programming equation for a drone to fly around multiple targets is:

[0125]

[0126] By calculating the optimal circling sequence of the drone, the redundancy problem of the flight path is optimized, the route length is shortened, and the working efficiency of the drone is improved. Figure 4 As shown, according to the shortest straight line between two points, the intersection of the line connecting the take-off point of the UAV and target 1 and the circle is taken as the first circling point of target 1, and then target 1 is circled in a clockwise or counterclockwise direction to obtain the last circling point of target 1, and the circling point is connected with target 2, and the intersection with the circumference of target 2 is taken as its first circling point, and then target 2 is circled in a clockwise or counterclockwise direction. If there are more targets, and so on, multiple targets are circled one by one to obtain their dynamic starting and ending circling points, which effectively optimizes the problem of flight path redundancy.

[0127] Step 203, the UAV collects data on all targets at different altitudes according to the optimal circling sequence for each target;

[0128] The process of the drone collecting data on various targets at different altitudes is as follows: for a certain target, the drone flies around at a rotation angle interval of α, and collects a photo of the target every time the drone rotates by an angle α until it circles the entire target and collects photos from multiple angles.

[0129] Step 3: Set the latitude and longitude coordinates of each target [lng i ,lat i ] is converted to plane coordinates [x i ,y i ], and calculate the plane coordinates of each waypoint corresponding to each target according to the flying radius and rotation angle of the drone, and then go to step five;

[0130] like Figure 2 As shown, the coordinates of each target are converted as follows: first, the latitude and longitude coordinates of the target are converted to WGS84 coordinates, and then the WGS84 coordinates are converted to plane coordinates.

[0131] Plane coordinates of each target corresponding to each waypoint It is expressed as:

[0132]

[0133]

[0134] Where j represents the jth waypoint number.

[0135] Step 4: The fixed-wing UAV equipped with synthetic aperture radar scans each target area of ​​the route planning to obtain the plane coordinates of the scanned area;

[0136] Specifically:

[0137] Step 401, inputting the coordinates of the center point of the scanning area corresponding to the current target, the width L1 and length L2 of the rectangular scanning area, the UAV flight altitude H', the ground rubbing angle β and the width W' of the UAV strip coverage area into the synthetic aperture radar;

[0138] like Figure 5 As shown, if the UAV performs multi-view route planning, the scanning area is a square, that is, L1 = L2;

[0139] Step 402: Based on the width L1 and length L2 of the rectangular scanning area, the flight height H', the ground rubbing angle β and the coordinates (X O ,Y O ), calculate the vertex coordinates of the rectangular scanning area, and obtain the quadrilateral flight route of the UAV.

[0140] If it is multi-perspective route planning, such as Figure 6 As shown, to ensure that the target area can be completely covered from all angles, a rectangle is made with the diagonal length of the square scanning area 7 / 5L1 outward, with the east direction as the positive direction of the X axis. The coordinates of the four vertices of the rectangle are:

[0141]

[0142]

[0143]

[0144]

[0145] If it is a full coverage route planning, the principle is as follows Figure 7 As shown in Figure 1, a full coverage scan is performed on an area to collect data. Figure 8 As shown, with the east direction as the positive direction of the X-axis, the coordinates of the four vertices of the rectangular scanning area are:

[0146]

[0147] X n =X1+(u-1)×F

[0148]

[0149]

[0150] Among them, F is the route spacing, and 500 is the length reserved for the start and end points of each route. u represents the uth route. The fixed-wing UAV equipped with synthetic aperture radar flies back and forth, and eventually there will be U routes in a rectangular area.

[0151] Step 5, converting the plane coordinates of each waypoint obtained in step 3 or step 4 into longitude and latitude coordinates, obtaining each route waypoint of the drone flying around the target, and generating a specific route file;

[0152] like Figure 2 As shown, the specific process of converting plane coordinates into longitude and latitude coordinates is: first convert the plane coordinates into WGS84 coordinates, and then convert the WGS84 coordinates into longitude and latitude coordinates.

[0153] Step 6: Import the route file into the flight control system of the drone, plan the actual flight route of the drone, control the drone to fly along the set route and collect data.

Claims

1. A route planning method based on multi-band camouflage feature data collection, characterized in that: The specific steps include: Step 1: According to the type of drone and the data collection equipment it carries, the multi-rotor drone equipped with a visible light or infrared camera among the drones for route planning shall execute step 2; the fixed-wing drone equipped with a synthetic aperture radar shall execute step 4; Step 2: Consider the impact of weather on multi-rotor drones equipped with visible light or infrared cameras, calculate the actual flight altitude of the drones for data collection on the targets, and plan the optimal order for the drones to fly around the targets; When the drone collects photos vertically on the ground, the drone's theoretical flight altitude H is calculated using the camera payload parameters, and the flight altitude is adjusted according to weather conditions to obtain the drone's actual flight altitude h; Due to the influence of weather, the resolution of the onboard camera for shooting the target is guaranteed by changing the flight altitude of the drone, specifically: ①Fog will weaken the reflected light of the target, and the light will be scattered into the camera by the fog. Then, under foggy weather conditions, the light intensity t(x) received by the drone's onboard camera is: t(x)=e -γd(x) Among them, d(x) is the depth of field, that is, the height of the lens from the ground when the drone collects data; γ is the scattering coefficient caused by fog; λ is the wavelength, in microns; Q is the visibility of fog, in meters; The light intensity t(x) reaching the airborne camera decays exponentially with the increase of the fog scattering coefficient γ. The flight altitude h of the drone is equivalent to the depth of field d(x), so the actual flight altitude of the drone is: h=ln(t(x)) / γ If the light intensity t(x) received by the camera is to be kept constant, the flight altitude of the drone is lowered to ensure the clarity of the captured image and the distinguishability of the target; ②The influence of humidity and carbon dioxide on the atmospheric transmittance of infrared cameras is: Infrared atmospheric transmittance = radiation flux after propagation attenuation / radiation flux before incident When the infrared wavelength is 8-12μm, the influence of water vapor and carbon dioxide on the atmospheric transmittance τ of infrared radiation is exponential, specifically: t H2O =0.9993×e -0.01504x (8-12μm) t CO2 =0.199×e -0.02107x +0.8006×e -0.00112x (8-12μm) When the atmospheric humidity is greater than 85%, the drone equipped with an infrared camera stops shooting the target; when the atmospheric humidity is less than 85%, the flight altitude of the drone is adjusted. The higher the humidity, the lower the actual flight altitude. ③At the same time, the atmospheric transmittance and ambient temperature affect the target measurement temperature, further affecting the data collection effect of the infrared camera on the target; The measured temperature of the target t c for: Where τ represents the atmospheric transmittance, represents the target surface emissivity, t0 represents the actual temperature of the target being measured, t a Indicates the ambient temperature; The greater the temperature difference between the ambient temperature and the actual temperature of the target, and the greater the target measurement temperature, the better the data collection effect on the target; Step 3: Set the latitude and longitude coordinates of each target [lng i ,lat i ] is converted to plane coordinates [x i ,y i ], and calculate the plane coordinates of each waypoint corresponding to each target according to the flying radius and rotation angle of the drone, and then go to step five; Step 4: The fixed-wing UAV equipped with synthetic aperture radar scans each target area of ​​the route planning to obtain the plane coordinates of the scanned area; Step 5, converting the plane coordinates of each waypoint obtained in step 3 or step 4 into longitude and latitude coordinates, obtaining each route waypoint of the drone flying around the target, and generating a specific route file; Step 6: Import the route file into the flight control system of the drone, plan the actual flight route of the drone, control the drone to fly along the set route and collect data.

2. A route planning method based on multi-band camouflage feature data collection according to claim 1, characterized in that: The theoretical flight height H of the multi-rotor drone equipped with a visible light or infrared camera is: The parameters of the camera payload include the focal length of the camera lens f, the pixel size of the camera a, and the ground resolution GSD of the photo.

3. A route planning method based on multi-band camouflage feature data collection according to claim 2, characterized in that: The multi-rotor drone equipped with a visible light or infrared camera obtains a circling radius for each target according to the actual flight altitude; Assume that the length and width of the target are L and W respectively, the distance between the camera lens and the target is S, the actual flight height of the drone perpendicular to the ground is h, the flight radius is r, the proportion of the target in the target photo taken by the camera is y, the resolution of the photo is p cm / pixel, and the photo size is m×n, then the radius r of the drone flying around the target is: The distance S between the camera lens and the target is: The photo ground resolution GSD is calculated based on the photo resolution, photo size, and target length and width:

4. A route planning method based on multi-band camouflage feature data collection according to claim 1, characterized in that: The specific process of determining the optimal order for the multi-rotor drone equipped with visible light or infrared cameras to fly around the target is as follows: Step 201, input the longitude and latitude of the take-off point of the drone and each target into the drone route planning program. The drone takes off from the take-off point s, and the final destination is i. According to the calculated actual flight altitude and circling radius, the drone circling each target, and calculates the shortest path length d(i,v) from the destination i, passing through each target in V once and only once, and finally returning to the take-off point s; V is the set of other target points except the target point i; The distance from the destination point i to any destination point k in the set V is expressed as c ik express; Step 202, determine whether the set V is an empty set. If so, d(i, v) means directly returning from the destination point i to the starting point s. At this time, d(i, v) = c is And i≠s, c is represents the distance from the destination point i to the starting point s; when the set V is not an empty set, the drone attempts to fly around each target in the set V and finds the shortest flying distance, which is the optimal order of flying around the targets; The optimal order for a drone to fly around multiple targets is to solve the equation: Step 203, the UAV collects data on all targets at different altitudes according to the optimal circling sequence for each target; The process of the drone collecting data on various targets at different altitudes is as follows: for a certain target, the drone flies around at a rotation angle interval of α, and collects a photo of the target every time the drone rotates by an angle α until it circles the entire target and collects photos from multiple angles.

5. The route planning method based on multi-band camouflage feature data collection according to claim 1 is characterized in that: The plane coordinates of each waypoint corresponding to each target described in step 3 It is expressed as: Among them, j represents the j-th waypoint number.

6. The route planning method based on multi-band camouflage feature data collection according to claim 1 is characterized in that: The calculation process of the plane coordinates of the scanning area by the fixed-wing UAV equipped with synthetic aperture radar is specifically as follows: Specifically: Step 401: input the coordinates of the center point of the scanning area corresponding to the current target, the width L1 and length L2 of the rectangular scanning area, the UAV flight altitude H', the ground grazing angle β, and the width W' of the UAV strip coverage area into the synthetic aperture radar; If the drone performs multi-view route planning, the scanning area is a square, that is, L1 = L2; Step 402: Based on the width L1 and length L2 of the rectangular scanning area, the flight height H', the ground rubbing angle β and the coordinates (X O ,Y O ), calculate the vertex coordinates of the rectangular scanning area, that is, obtain the quadrilateral flight route of the UAV; If it is a multi-view route planning, a rectangle is made with the diagonal length 7 / 5L1 of the square scanning area, with the east direction as the positive direction of the X axis. The coordinates of the four vertices of the rectangle are: If the route is planned for full coverage, with the east direction as the positive direction of the X axis, the coordinates of the four vertices of the rectangular scanning area are: X n =X1+(u-1)×F Among them, F is the route spacing, 500 is the length reserved for the starting point and end point of each route; u represents the u-th route. The fixed-wing UAV equipped with synthetic aperture radar flies back and forth, and eventually there will be U routes in a rectangular area.

Images