Unmanned aerial vehicle target infrared image simulation method based on rough surface BRDF model

Abstract

The invention discloses an unmanned aerial vehicle target infrared image simulation method based on a rough surface BRDF model. The method comprises the following steps: obtaining an unmanned aerial vehicle three-dimensional model; calculating the temperature distribution of the unmanned aerial vehicle by using ANSYS software; introducing surface roughness into the unmanned aerial vehicle three-dimensional model, and establishing an unmanned aerial vehicle surface light intensity distribution model based on a BRDF model; based on an unmanned aerial vehicle temperature distribution and surface light intensity distribution model, establishing an unmanned aerial vehicle target infrared radiation characteristic model; and generating an unmanned aerial vehicle target infrared simulation image based on the detector imaging projection model. The infrared radiation characteristic of the unmanned aerial vehicle generated by image simulation conforms to the infrared radiation characteristic of the actually shot unmanned aerial vehicle.

Description

Technical Field

[0001] The invention relates to the technical field of infrared image simulation, and in particular to an infrared image simulation method of a UAV target based on a rough surface BRDF model. Background Art

[0002] In recent years, the UAV industry has developed rapidly. Due to its low cost and the lack of effective detection and countermeasures in the industry, the phenomenon of "illegal flying" of UAVs has become increasingly serious. At present, infrared thermal imaging detection is an important means of identifying UAVs, so the study of the infrared radiation target characteristics of UAVs is of great significance. Real shooting in outdoor scenes is an effective means to obtain real UAV infrared images, but this method consumes a lot of time and manpower and material resources to obtain infrared images under specific environmental parameters, so the efficiency is relatively low. In order to improve the efficiency of obtaining UAV infrared images, the use of simulation methods to obtain infrared images has received more and more attention.

[0003] At present, most of the work on aircraft infrared image simulation is focused on the study of aircraft targets. The tail flame, nozzle and other parts of the aircraft have strong heat sources, but the infrared radiation of the quadrotor drone target is weak. The infrared characteristics and infrared image simulation of small targets at long distances mostly approximate them to point source targets, ignoring the distribution of heat sources, and there is little research on the simulation methods of quadrotor drones.

[0004] Jia Qinglian established a theoretical calculation model for the skin, tail nozzle and tail flame radiation of the UAV in the non-afterburner state, and used the finite element analysis method to solve the infrared radiation calculation formula at different azimuths and different observation elevation angles. Liu Lianwei and others modeled and simulated the infrared radiation characteristics of a typical large aspect ratio propeller UAV, calculated the temperature distribution of the UAV using CFD software, established a steady-state temperature database for the skin and a temperature interpolation algorithm that met the accuracy requirements, and realized the interpolation calculation of the UAV skin temperature.

[0005] In summary, the infrared radiation characteristics of aircraft with strong heat sources, such as airplanes and gasoline-powered drones, have been widely studied and gradually improved. Unfortunately, there are relatively few studies on the infrared radiation characteristics of civilian low-speed, small and slow drones, and most of them are obtained through direct experimental measurements. Summary of the invention

[0006] In order to overcome the above-mentioned deficiencies of the prior art, the present invention provides a UAV target infrared image simulation method based on a rough surface BRDF model to improve the authenticity and acquisition efficiency of infrared images of low, small and slow UAVs.

[0007] The technical solution to achieve the purpose of the present invention is: a method for simulating infrared images of unmanned aerial vehicle targets based on a rough surface BRDF model, comprising the following steps:

[0008] Step 1: Obtain a three-dimensional model of a drone, wherein the three-dimensional model of the drone includes a drone wing, a motor, a fuselage, and a battery;

[0009] Step 2: Calculate the temperature distribution of the drone itself, use ANSYS software to define the model material and related parameters and divide the grid, obtain the temperature distribution of the drone itself, and distribute the target temperature into the simulation system, reconstruct and simulate the node temperature and cloud point coordinates;

[0010] Step 3: Construct a light intensity distribution model of the drone surface, introduce surface roughness into the drone 3D model, and establish a light intensity distribution model of the drone surface based on the BRDF model;

[0011] Step 4: Based on the temperature distribution and surface light intensity distribution model of the UAV itself, establish the infrared radiation characteristic model of the UAV target;

[0012] Step 5: Generate an infrared simulation image of the UAV target based on the detector imaging projection model.

[0013] Preferably, the specific method of step 2 is:

[0014] Step 2.1, establish steady-state and transient thermal systems in Ansys Workbench and define the materials required for model calculation;

[0015] Step 2.2: Pre-process and repair the UAV 3D model in SpaceClaim, delete the parts that have no effect on heat transfer, divide the entities and assign materials to them;

[0016] Step 2.3, import the pre-processed model into Ansys Mechanical, divide the mesh, define the mesh parameters, and obtain the mesh nodes and cloud point coordinates of the model after the mesh division is completed;

[0017] Step 2.4: Analyze the heat source of the drone itself to obtain the heat transfer boundary conditions, and solve the temperature distribution of the drone itself in steady state and transient state through Ansys Workbench;

[0018] Step 2.5, export the node coordinates and temperature data files in txt format, and export the 3D model file after meshing in stl format;

[0019] Step 2.6, import the spatial coordinates of the grid nodes divided in step 2.3 and the temperature data corresponding to the node coordinates into the simulation program;

[0020] Step 2.7, read the three-dimensional model file exported in step 2.3, obtain the spatial topological relationship of the nodes and filter out the nodes that do not correspond to the model grid, connect the nodes of the same microfacet, and thus reconstruct the geometric model of the target in the simulation system;

[0021] Step 2.8: Attach the temperature data corresponding to the node coordinates to the reconstructed geometric model to obtain the temperature field distribution of the target on the simulation platform.

[0022] Preferably, in step 3, the light intensity distribution on the surface of the drone is constructed, the surface roughness is introduced into the three-dimensional model of the drone, and the specific steps of establishing the light intensity distribution model on the surface of the drone based on the BRDF model are as follows:

[0023] Step 3.1, read the microfacet coordinate data obtained in step 2.3, and calculate the area of ​​each microfacet;

[0024] Step 3.2, determine whether the surface element radiation can be received by the detector. If it can be received, further determine whether the sunlight can irradiate the surface element. If it cannot be received by the detector, calculate the next surface element until all the surface elements are judged;

[0025] Step 3.3, calculate the distribution of ambient radiation reflected from the surface of the drone;

[0026] Step 3.4: Calculate the solar radiation and sky radiation received by each facet.

[0027] Preferably, the specific method for determining whether the surface element radiation can be received by the detector is: calculating the outward normal vector of each grid micro-surface element; calculating whether the angle between the micro-surface element normal vector and the outward normal vector of the detector pixel is an acute angle, if it is an acute angle, the surface element radiation can be received by the detector.

[0028] Preferably, the specific method for determining whether sunlight can irradiate the surface element is:

[0029] Determine whether the angle between the sunlight and the microfacet normal vector is an acute angle. If it is an acute angle, it means that the sunlight can irradiate the microfacet.

[0030] Preferably, the specific formula for calculating the distribution of ambient radiation reflected from the surface of the drone is:

[0031]

[0032] Where F is the Fresnel reflectivity, D is the microsurface normal distribution function, and G is the geometric attenuation factor. is the normal direction of the microfacet; is the direction vector of the incident light; is the viewing direction vector, R sDistribution function of ambient radiation reflected from the drone surface.

[0033] Preferably, the calculation formulas for solar radiation illuminance and sky radiation illuminance are:

[0034] P sun =Q sun ×A×R S

[0035]

[0036] Among them, P sun is the solar radiation received by the surface element, Q sun is the solar radiation emittance set for simulation, A is the microfacet area, P is sky is the sky radiation received by the surface element, Q sky Sky background radiance emittance set for the simulation.

[0037] Preferably, in step 4, the specific method for establishing the infrared radiation characteristic model of the drone target based on the drone's own temperature distribution and surface light intensity distribution model is:

[0038] Step 4.1, according to the temperature of the node on each unit and the material properties of the unit, obtain the infrared radiation distribution of each unit;

[0039] Step 4.2: Superimpose the infrared radiation of each unit with the reflected external radiation to obtain the overall infrared radiation distribution of the drone.

[0040] Compared with the prior art, the present invention has the following significant advantages: (1) The present invention uses Ansys software to calculate the temperature distribution of the drone itself to improve the authenticity of the drone simulation. (2) The present invention uses a rough surface BRDF model to calculate the impact of solar radiation and sky background radiation on drone radiation, which better simulates the radiation characteristic distribution of a real drone in the sky. By changing the incident light source position and the detection direction and distance, a series of drone infrared simulation images with different light source incident angles and detection directions and distances can be obtained, enriching the drone infrared image data set. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 The present invention is a flowchart of a method for simulating infrared images of UAV targets based on a rough surface BRDF model.

[0042] Figure 2 Schematic diagram of the process of projecting the radiation illumination of the micro-element surface to the detector pixel.

[0043] Figure 3 This is a 3D model diagram of a drone.

[0044] Figure 4 Calculate the drone's own temperature map for Ansys simulation.

[0045] Figure 5 Schematic diagram of the BRDF model used, which is the Cook-Torrance model.

[0046] Figure 6 A series of infrared simulation images of drones with different light source incident angles, detection directions and distances generated by simulation. Specific implementation plan

[0047] A UAV target infrared image simulation method based on a rough surface BRDF model is proposed to improve the authenticity and acquisition efficiency of infrared images of low, small and slow UAVs.

[0048] In order to make the simulated drone more realistic, the present invention makes improvements in three aspects: (1) Select a suitable drone 3D model file. The selected model adopts multi-entity modeling method when using SolidWorks modeling, and is exported to Parasolid format, i.e. x_t format. The model includes drone wings, motors, fuselage, batteries and other heat sources. In this way, the model can be directly generated into ANSYS without manual splicing and Boolean operations, which improves the efficiency of target geometry modeling. (2) Use ANSYS to calculate the temperature field distribution of the drone itself. ANSYS has mature and rich flow field and heat transfer modules, which ensures accuracy while avoiding the need to supplement the calculation of flow field and heat transfer formulas in the algorithm, thereby improving the calculation efficiency. (3) Use the rough surface BRDF model to calculate the influence of solar radiation and sky background radiation on drone radiation, which better simulates the radiation characteristic distribution of real drones in the sky.

[0049] The present invention proposes a method for simulating infrared images of unmanned aerial vehicle targets based on a rough surface BRDF model, comprising the following steps:

[0050] Step 1: Get the 3D model of the drone, use SolidWorks to model it, and export it to Parasolid format, i.e. x_t format. The model includes the drone wings, motors, fuselage, and heat sources such as batteries. This way, the model can be directly generated into ANSYS without manual splicing and Boolean operations, which improves the efficiency of target geometry modeling. Figure 3 The prototype of the drone 3D model used is the DJI Phantom 3 drone.

[0051] Step 2: Calculate the temperature distribution of the drone itself. Use ANSYS software to define the model material and related parameters and divide the grid to obtain the temperature distribution of the drone itself. Then, the temperature distribution of the drone target itself calculated by ANSYS is imported into the simulation system to reconstruct and simulate the node temperature and cloud point coordinates.

[0052] Step 2.1. Establish a steady-state thermal system in Ansys Workbench and define the materials required for model calculation, including structural steel, magnesium alloy, aluminum alloy, carbon fiber, nylon and other materials. Since Ansys has a rich engineering database, the selected material contains the parameter information required for heat transfer, mainly thermal conductivity and density.

[0053] Step 2.2: Pre-process the 3D model of the drone in SpaceClaim, delete the camera part and frame decoration part that have little effect on heat transfer, delete the extra facets of the fuselage during modeling, divide the propeller into one entity, divide the motor and rotor as a whole into one entity, divide the fuselage and frame into one entity, and divide the battery into one entity; perform model inspection and model repair on the divided entities;

[0054] Step 2.3, meshing, import the pre-processed model into Ansys Mechanical, define mesh properties, use adaptive adjustment, default resolution, feature removal, fast transition, large scale across angle center, medium smoothness, and the rest are default. After meshing is completed, the mesh nodes of the model and the cloud point coordinates corresponding to the mesh nodes can be obtained.

[0055] Step 2.4: Analyze the heat source of the drone to obtain the heat transfer boundary conditions, define the material of each entity part, and after the material is defined, the heat transfer properties are also defined. Define the initial temperature, set the radiation relationship between micro-facets, set the convection coefficient, and if it is transient, set the time parameter. After the above settings are completed, use Ansys Mechanical to solve the temperature distribution of the drone itself in steady state and transient state, such as Figure 4 As shown;

[0056] Step 2.5, export the node coordinates and temperature data file in txt format, and export the 3D model file after meshing in stl format.

[0057] Step 2.6, import the spatial coordinates of the grid nodes divided in step 2.3 and the temperature data corresponding to the node coordinates in step 2.4 into the simulation program;

[0058] Step 2.7, read the three-dimensional model file exported in step 2.3, obtain the spatial topological relationship of the nodes and filter out the nodes that do not correspond to the model grid, connect the nodes of the same microfacet, and thus reconstruct the geometric model of the target in the simulation system;

[0059] Step 2.8. Finally, the temperature data corresponding to the node coordinates are attached to the reconstructed geometric model, thereby obtaining the temperature field distribution of the target on the simulation platform.

[0060] Step 3: Construct the light intensity distribution of the drone surface, introduce surface roughness into the drone 3D model, and establish a drone surface light intensity distribution model based on the BRDF model;

[0061] Step 3.1, read the microfacet coordinate data obtained in step 2.3, and calculate the area of ​​each microfacet. The smaller the average area of ​​the microfacet, the better the quality of the simulated image, but the greater the amount of calculation, the slower the program runs; the larger the average area of ​​the microfacet, the faster the program runs, but the quality of the simulated image cannot be guaranteed, and striped erroneous images may appear. Therefore, it is necessary to select a suitable microfacet area as the segmentation threshold to increase the software running speed without affecting the quality of the simulated image. If the area of ​​the microfacet is higher than the threshold condition, the segmentation will continue until the microfacet area is less than the threshold.

[0062] Step 3.2, determine whether the surface element radiation can be received by the detector. If it can be received, further determine whether the sunlight can shine on the surface element. If it cannot be received by the detector, calculate the next surface element until all surface elements are judged. Judgment method: Calculate the outward normal vector of each grid micro-surface element, and calculate whether the angle between the micro-surface element normal vector and the outward normal vector of the detector pixel is an acute angle. If it is an acute angle, the surface element radiation can be received by the detector. The judgment of sunlight exposure is similar. The program assumes that the sun is shining from a certain position, and calculates whether the angle between the sunlight and the micro-surface element normal vector is an acute angle. If it is an acute angle, it means that the sunlight can shine on the micro-surface element. Figure 2 shown.

[0063] Step 3.3, calculate the distribution of ambient radiation reflected from the surface of the drone. The present invention adopts a bidirectional reflectance distribution function BRDF model, which can more accurately describe the reflection distribution of the object surface. The BRDF model adopted in the present invention is the Cook-Torrance model. The Cook-Torrance model is a more realistic calculation model proposed on the basis of the Torrance-Sparrow model. Its mirror reflectivity is a function of wavelength, which makes up for the deficiency of the Torrance-Sparrow model that cannot distinguish between targets of different materials and light sources of different properties. It points out that only the microplanes in the midline direction of the line of sight and the reflection direction contribute to the surface reflection term, which simplifies the model and increases the speed of calculation, and is suitable for real-time simulation. The geometric model of the Cook-Torrance model is as follows: Figure 5 As shown:

[0064] in, is the normal direction of the microfacet; is the direction vector of the incident light; is the viewing direction vector; The incident direction and viewing direction The half-angle direction vector of .

[0065] The specular reflection formula of the Cook-Torrance lighting model is as follows:

[0066]

[0067] Where F is the Fresnel reflectivity, which depends on the properties of the surface material and is often approximated in real-time 3D graphics, namely:

[0068]

[0069] Among them, n 1 and n 2 are the refractive indices of the two media, θ is the refraction angle,

[0070] D is the microsurface normal distribution function, usually a Gaussian distribution:

[0071]

[0072] Among them, m reflects the roughness of the surface material.

[0073] G is the geometric attenuation factor, which mainly considers the attenuation of the specular reflection component caused by the occlusion between micro-facets, which prevents some light from being reflected:

[0074]

[0075] Step 3.4: Calculate the solar radiation and sky radiation received by each facet using the formula in step 3.3:

[0076] P sun =Q sun ×A×R S

[0077]

[0078] Among them, P sun is the solar radiation received by the surface element, Q sun is the solar radiation emittance set for simulation, A is the microfacet area, P is sky is the sky radiation received by the surface element, Q sky Sky background radiance emittance set for the simulation.

[0079] Step 4: Based on the temperature distribution and surface light intensity distribution model of the UAV itself, establish the infrared radiation characteristic model of the UAV target;

[0080] Step 4.1: According to the temperature of the nodes on each unit and the material properties of the unit, the infrared radiation distribution of each unit is obtained.

[0081] The radiation formula used for self-radiation is the Stefan-Boltzmann law:

[0082] P self =ε×σ×T 4

[0083] Among them, P self is the radiation of the drone itself, ε is the emissivity of the material itself, σ is the Stefan-Boltzmann constant, and T is the node temperature.

[0084] Step 4.2: Superimpose each unit’s own infrared radiation and reflected external radiation to obtain the overall infrared radiation distribution of the drone.

[0085] P=P sun +P sky +P self

[0086] Step 5: Generate an infrared simulation image of the drone target based on the detector imaging projection model. The detector imaging projection model includes the camera matrix, light source parameters, and imaging parameters. By calculating the imaging position, the target's own radiation intensity, and the reflected external radiation intensity, the image grayscale value of the model is obtained, that is, the simulation image of the drone target.

[0087] Step 5.1, define the camera parameter matrix, including the camera intrinsic matrix and the camera extrinsic matrix. The intrinsic matrix includes the focal length, the coordinates of the principal point (optical center), and the correction coefficient. The extrinsic matrix includes the parameters of the position and posture of the camera in the world coordinate system.

[0088] Step 5.2, define the light source parameters, namely the incident zenith angle and azimuth angle of the light source, the camera detection zenith angle and azimuth angle, and the detection distance from the target; define the imaging parameters, including the imaging size, set the length to 200 and the width to 200; the image resolution, namely the number of image pixels.

[0089] Step 5.3, project the three-dimensional coordinate point to the two-dimensional coordinate plane, calculate the micro-facet projection to the pixel position, determine whether the radiation light of the drone facet can be received by the detector pixel (similar to step 3.2), calculate the radiation intensity of the projection according to step 4.2, and superimpose all the radiation light intensities received by the pixel to obtain the grayscale value of the pixel.

[0090] Step 5.4: Calculate the grayscale values ​​of all pixels in the image to obtain a simulated image of the UAV target.

[0091] The infrared simulation image results are as follows Figure 6As shown, (a)(b)(c) are simulation diagrams of different detection angles, (d)(e)(f) are simulation diagrams of different illumination directions, and (g)(h)(i) are simulation diagrams of different detection distances.

[0092] The temperature distribution of the drone itself calculated based on ANSYS in the present invention can better solve the heat transfer problem of different heat sources of the drone itself; the drone surface temperature calculation equation established by the present invention based on the BRDF model considering comprehensive influencing factors such as solar radiation, sky background radiation, and atmospheric convection can more accurately solve the absorption of external radiation by micro-facets on the drone surface; based on the imaging projection model, the present invention projects the radiation imaging of the drone surface onto the detector to form a radiation illumination image of the detector pixel, thereby realizing a rapid imaging method for drone targets.

Claims

1. A method for simulating infrared images of unmanned aerial vehicle targets based on a rough surface BRDF model, characterized in that: The following steps are involved: Step 1: Obtain a three-dimensional model of a drone, wherein the three-dimensional model of the drone includes a drone wing, a motor, a fuselage, and a battery; Step 2: Calculate the temperature distribution of the drone itself, use ANSYS software to define the model material and related parameters and divide the grid, obtain the temperature distribution of the drone itself, and distribute the target temperature into the simulation system, reconstruct and simulate the node temperature and cloud point coordinates; Step 3: Construct a light intensity distribution model of the drone surface, introduce surface roughness into the drone 3D model, and establish a light intensity distribution model of the drone surface based on the BRDF model; Step 4: Based on the temperature distribution and surface light intensity distribution model of the UAV itself, establish the infrared radiation characteristic model of the UAV target; Step 5: Generate an infrared simulation image of the UAV target based on the detector imaging projection model.

2. The infrared image simulation method of a UAV target based on a rough surface BRDF model according to claim 1 is characterized in that: The specific method of step 2 is: Step 2.1, establish steady-state and transient thermal systems in Ansys Workbench and define the materials required for model calculation; Step 2.2: Pre-process and repair the UAV 3D model in SpaceClaim, delete the parts that have no effect on heat transfer, divide the entities and assign materials to them; Step 2.3, import the pre-processed model into Ansys Mechanical, divide the mesh, define the mesh parameters, and obtain the mesh nodes and cloud point coordinates of the model after the mesh division is completed; Step 2.4: Analyze the heat source of the drone itself to obtain the heat transfer boundary conditions, and solve the temperature distribution of the drone itself in steady state and transient state through Ansys Workbench; Step 2.5, export the node coordinates and temperature data files in txt format, and export the 3D model file after meshing in stl format; Step 2.6, import the spatial coordinates of the grid nodes divided in step 2.3 and the temperature data corresponding to the node coordinates into the simulation program; Step 2.7, read the three-dimensional model file exported in step 2.3, obtain the spatial topological relationship of the nodes and filter out the nodes that do not correspond to the model grid, connect the nodes of the same microfacet, and thus reconstruct the geometric model of the target in the simulation system; Step 2.8: Attach the temperature data corresponding to the node coordinates to the reconstructed geometric model to obtain the temperature field distribution of the target on the simulation platform.

3. The infrared image simulation method of a UAV target based on a rough surface BRDF model according to claim 1 is characterized in that: In step 3, the light intensity distribution of the drone surface is constructed, and the surface roughness is introduced into the drone 3D model. The specific steps to establish the drone surface light intensity distribution model based on the BRDF model are as follows: Step 3.1, read the microfacet coordinate data obtained in step 2.3, and calculate the area of ​​each microfacet; Step 3.2, determine whether the surface element radiation can be received by the detector. If it can be received, further determine whether the sunlight can irradiate the surface element. If it cannot be received by the detector, calculate the next surface element until all the surface elements are judged; Step 3.3, calculate the distribution of ambient radiation reflected from the surface of the drone; Step 3.4: Calculate the solar radiation and sky radiation received by each facet.

4. The infrared image simulation method of a UAV target based on a rough surface BRDF model according to claim 3 is characterized in that: The specific method for determining whether the surface element radiation can be received by the detector is as follows: calculating the outward normal vector of each grid micro-surface element; Calculate whether the angle between the micro-surface element normal vector and the outward normal vector of the detector pixel direction is an acute angle. If it is an acute angle, the surface element radiation energy can be received by the detector.

5. The method for simulating infrared images of unmanned aerial vehicle targets based on a rough surface BRDF model according to claim 3, characterized in that: The specific method for judging whether sunlight can irradiate the surface element is: Determine whether the angle between the sunlight and the microfacet normal vector is an acute angle. If it is an acute angle, it means that the sunlight can irradiate the microfacet.

6. The method for simulating infrared images of unmanned aerial vehicle targets based on a rough surface BRDF model according to claim 3, characterized in that: The specific formula for calculating the distribution of ambient radiation reflected from the surface of the drone is: Where F is the Fresnel reflectivity, D is the microsurface normal distribution function, and G is the geometric attenuation factor. is the normal direction of the microfacet; is the direction vector of the incident light; is the viewing direction vector, R s Distribution function of ambient radiation reflected from the drone surface.

7. The method for simulating infrared images of unmanned aerial vehicle targets based on a rough surface BRDF model according to claim 3, characterized in that: The calculation formulas for solar radiation and sky radiation are: P sun =Q sun ×A×R S Among them, P sun is the solar radiation received by the surface element, Q sun is the solar radiation emittance set for simulation, A is the microfacet area, P is sky is the sky radiation received by the surface element, Q sky Sky background radiance emittance set for the simulation.

8. The method for simulating infrared images of unmanned aerial vehicle targets based on a rough surface BRDF model according to claim 1, characterized in that: In step 4, the specific method for establishing the infrared radiation characteristic model of the drone target based on the drone's own temperature distribution and surface light intensity distribution model is as follows: Step 4.1, according to the temperature of the node on each unit and the material properties of the unit, obtain the infrared radiation distribution of each unit; Step 4.2: Superimpose the infrared radiation of each unit with the reflected external radiation to obtain the overall infrared radiation distribution of the drone.

9. The method for simulating infrared images of unmanned aerial vehicle targets based on a rough surface BRDF model according to claim 8, characterized in that: The infrared radiation distribution of each unit is obtained as follows: P self =ε×σ×T 4 Among them, P self is the infrared radiation of each unit of the drone, ε is the emissivity of the material itself, σ is the Stefan-Boltzmann constant, and T is the node temperature.

10. The method for simulating infrared images of unmanned aerial vehicle targets based on a rough surface BRDF model according to claim 1, characterized in that: Step 5 Based on the detector imaging projection model, the specific steps of generating the infrared simulation image of the UAV target are as follows: Step 5.1, define a camera parameter matrix, wherein the camera parameter matrix includes a camera intrinsic parameter matrix and a camera extrinsic parameter matrix; Step 5.2, define the light source parameters, i.e., the incident zenith angle and azimuth angle of the light source, the camera detection zenith angle and azimuth angle, and the target detection distance; define the imaging parameters, including the imaging size and image resolution, i.e., the number of image pixels; Step 5.3, project the three-dimensional coordinate point onto the two-dimensional coordinate plane, calculate the micro-facet projection onto the pixel position, determine whether the radiation light of the drone facet can be received by the detector pixel, and superimpose all the radiation light intensities received by the pixel according to the projected radiation intensity to obtain the gray value of the pixel; Step 5.4: Calculate the grayscale values ​​of all pixels in the image to obtain a simulated image of the UAV target.

Images