Unmanned aerial vehicle three-dimensional environment modeling method and system based on millimeter wave radar

Abstract

The invention provides an unmanned aerial vehicle three-dimensional environment modeling method and system based on a millimeter-wave radar. The method comprises the following steps: transmitting an optimized millimeter-wave signal to a preset vegetation area by using the millimeter-wave radar to obtain a vegetation area reflection signal; carrying out inverse deduction on the signal propagation path of the vegetation area reflection signal, eliminating waveform distortion characteristics in the vegetation area reflection signal, and generating a distortion-eliminated reflection signal; fusing the reflected signal after distortion elimination with vegetation material features in the reflected signal of the vegetation area to generate a preset three-dimensional structure feature of the vegetation area; performing spatial reconstruction on the three-dimensional structural features based on spatial distribution information in the vegetation region reflection signals to generate a preset three-dimensional spatial model of the vegetation region; according to the method, direct physical conversion from the millimeter wave radar point cloud data to the three-dimensional model is realized, the problems of model cavity and material confusion caused by vegetation signal interference in traditional modeling are solved, and the scene sensing precision of the unmanned aerial vehicle in a dense vegetation area is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional environment modeling, and in particular to a method and system for three-dimensional environment modeling of an unmanned aerial vehicle (UAV) based on millimeter-wave radar. Background Art

[0002] With the widespread application of drones in forest resource monitoring, disaster warning, and ecological research, the demand for high-precision three-dimensional modeling of complex forest environments is becoming increasingly urgent. Therefore, there is an urgent need for a technical means that can still have good penetration ability and positioning accuracy under complex vegetation conditions to achieve efficient and detailed three-dimensional modeling of forest environments.

[0003] The current mainstream approach is to perform three-dimensional modeling of forest areas based on lidar combined with multi-view stereo vision technology. This involves using drones equipped with lidar systems to emit laser pulses, receive echo signals from the ground and vegetation, and then use point cloud fusion algorithms to construct a three-dimensional model with geometric structure and surface features, combined with texture information provided by high-resolution optical images. Existing approaches have several inherent flaws. These include the easy obstruction of radar signals in densely populated or complex forest environments, resulting in missing data on underlying vegetation structures; multipath reflections can cause ranging errors, affecting the overall accuracy of the model; and optical images rely on lighting conditions, making it challenging to obtain high-quality texture information in the low-light environment of the forest floor. This limits the authenticity and integrity of the modeling results, restricting the applicability and stability of existing technologies in complex forest scenarios. Summary of the Invention

[0004] The present invention provides a method and system for three-dimensional environment modeling of unmanned aerial vehicles based on millimeter-wave radar, which is used to solve the problems in the existing technology: radar signals are easily blocked in forest environments with high vegetation density or complex layers, resulting in data missing of the underlying vegetation structure; multipath reflection phenomena may cause ranging errors, affecting the overall accuracy of the model; optical images rely on lighting conditions, and obtaining high-quality texture information in the low-light environment under the forest is challenging, which limits the authenticity and integrity of the modeling results and restricts the applicability and stability of the existing technology in complex forest scenes.

[0005] In a first aspect, the present invention provides a method for modeling a three-dimensional environment of an unmanned aerial vehicle (UAV) based on millimeter-wave radar, comprising:

[0006] The millimeter-wave radar pre-deployed in the drone is used to transmit the optimized millimeter-wave signal into the preset vegetation area to obtain the reflected signal of the vegetation area;

[0007] Reversely inferring a signal propagation path of the vegetation area reflection signal, eliminating waveform distortion features in the vegetation area reflection signal, and generating a distortion-eliminated reflection signal;

[0008] fusing the distortion-eliminated reflection signal with vegetation material features in the vegetation area reflection signal to generate three-dimensional structural features of the preset vegetation area;

[0009] The three-dimensional structural features are spatially reconstructed based on the spatial distribution information in the reflection signal of the vegetation area to generate a three-dimensional spatial model of the preset vegetation area.

[0010] Optionally, a millimeter-wave radar pre-deployed in a UAV is used to transmit an optimized millimeter-wave signal into a preset vegetation area to obtain a reflected signal from the vegetation area, including:

[0011] Adjusting the impedance matching network parameters in a pre-deployed millimeter-wave radar to change the signal transmission characteristics of the transmitting circuit in the millimeter-wave radar and generate optimized impedance parameters;

[0012] Driving preset multiple frequency band millimeter wave signal sources based on the optimized impedance parameters to generate a combined frequency signal;

[0013] performing waveform shaping processing on the combined frequency signal to compress the pulse width of the combined frequency signal to generate an optimized millimeter wave signal;

[0014] The optimized millimeter wave signal is directionally radiated into a preset vegetation area to trigger a vegetation reflection echo, and the signal of the vegetation reflection echo is amplified to generate a vegetation area reflection signal.

[0015] Optionally, inversely calculating a signal propagation path of the vegetation area reflection signal, eliminating waveform distortion features in the vegetation area reflection signal, and generating a distortion-eliminated reflection signal includes:

[0016] Calculating the phase change rate of the phase time series in the vegetation area reflection signal to generate signal phase change information;

[0017] Calculating phase differences between adjacent receiving units in the millimeter-wave radar based on the signal phase change information to invert the electromagnetic wave propagation direction of the optimized millimeter-wave signal and generate a direct wave propagation path;

[0018] Reconstructing a preset wavefront energy distribution according to the direct wave propagation path to generate a direct wave component and a reflected wave component;

[0019] The energy intensity difference between the direct wave component and the reflected wave component is compared, and the interference beam in the reflected wave component is eliminated to generate a distortion-eliminated reflected signal.

[0020] Optionally, reconstructing a preset wavefront energy distribution according to the direct wave propagation path to generate a direct wave component and a reflected wave component includes:

[0021] Calculating a preset radius of wavefront curvature based on the direct wave propagation path to generate a wavefront curvature parameter;

[0022] Mapping the spatial energy gradient of the wavefront curvature parameter to generate a reconstructed wavefront energy field;

[0023] Scanning a preset energy continuity boundary condition in the reconstructed wavefront energy field to generate an energy mutation location;

[0024] The reconstructed wavefront energy field is divided into continuous energy regions and discontinuous energy regions based on the energy mutation position to generate a direct wave component and a reflected wave component.

[0025] Optionally, fusing the distortion-eliminated reflection signal with vegetation material features in the vegetation area reflection signal to generate the three-dimensional structural features of the preset vegetation area includes:

[0026] Analyzing the distance and azimuth information in the distortion-eliminated reflected signal to generate a target position coordinate set;

[0027] Matching vegetation material features in the reflection signal of the vegetation area with a preset material structure rule library to generate material structure attributes;

[0028] Constructing a spatial topological connection relationship between each position coordinate in the target position coordinate set based on the target position coordinate set and the material structure attribute;

[0029] The connection nodes in the spatial topological connection relationship are traversed to synthesize three-dimensional geometric structure units and generate three-dimensional structural features.

[0030] Optionally, matching vegetation material features in the vegetation area reflection signal with a preset material structure rule library to generate material structure attributes includes:

[0031] Extracting material electrical property parameters of vegetation material characteristics from the vegetation area reflection signal, converting the material electrical property parameters into a standardized identifier, and generating a material electrical identifier;

[0032] Matching the material electrical identification with a preset material threshold interval to determine a material type identification of the material electrical identification;

[0033] Based on the material type identifier, a query is performed in a preset plant structure rule library to obtain a structural connection rule corresponding to the material type identifier;

[0034] The material type identifier is combined with the structural connection rule to generate material structural attributes.

[0035] Optionally, spatially reconstructing the three-dimensional structural features based on spatial distribution information in the reflection signal of the vegetation area to generate a three-dimensional spatial model of the preset vegetation area includes:

[0036] parsing a three-dimensional coordinate point set of spatial distribution information in the reflection signal of the vegetation area to generate a spatial position topology map;

[0037] Mapping the three-dimensional geometric structure units in the three-dimensional structure features to the position nodes in the spatial position topology graph to generate a node mapping relationship;

[0038] Adjusting the spatial orientation of the three-dimensional geometric structure unit according to the node mapping relationship to generate an orientation-corrected structure unit;

[0039] All the orientation correction structural units are combined to generate a three-dimensional spatial model of the preset vegetation area.

[0040] In a second aspect, the present invention provides a millimeter-wave radar-based three-dimensional environment modeling system for unmanned aerial vehicles, comprising:

[0041] The transmitting module is used to transmit the optimized millimeter wave signal to the preset vegetation area using the millimeter wave radar pre-deployed in the UAV to obtain the reflected signal of the vegetation area;

[0042] an elimination module, configured to reversely infer a signal propagation path of the vegetation area reflection signal, eliminate waveform distortion features in the vegetation area reflection signal, and generate a distortion-eliminated reflection signal;

[0043] a fusion module, configured to fuse the distortion-eliminated reflection signal with vegetation material features in the vegetation area reflection signal to generate three-dimensional structural features of the preset vegetation area;

[0044] A reconstruction module is used to spatially reconstruct the three-dimensional structural features based on the spatial distribution information in the reflection signal of the vegetation area to generate a three-dimensional spatial model of the preset vegetation area.

[0045] In a third aspect, the present invention provides a computing device comprising a processor and a memory, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute a method for three-dimensional environment modeling of a drone based on millimeter-wave radar as described in any one of the first aspects.

[0046] In a fourth aspect, the present invention provides a computer storage medium having computer program instructions stored thereon, which, when executed by a processor, implements a method for three-dimensional environment modeling of a UAV based on millimeter-wave radar as described in any one of the first aspects.

[0047] The present invention eliminates waveform distortion caused by multipath reflection from vegetation by adopting the signal propagation path inversion technology, combines the material characteristics and geometric structure characteristics of the reflected signal for fusion processing, and then spatially reconstructs the three-dimensional structural characteristics based on the spatial distribution information. This method realizes the direct physical conversion of millimeter-wave radar point cloud data into a three-dimensional model in a forest environment, overcoming the model voids and material confusion problems caused by vegetation signal interference in traditional modeling, significantly improving the scene perception accuracy of drones in dense vegetation areas, and providing centimeter-level precision spatial model support for applications such as forest resource surveys and disaster monitoring.

[0048] Furthermore, by dynamically adjusting the impedance matching network of the millimeter-wave radar to optimize signal transmission efficiency, combining multi-band millimeter-wave signals to enhance penetration capability, and compressing the pulse width to improve distance resolution, and combining directional radiation with signal amplification technology, this method achieves stable acquisition of high signal-to-noise ratio reflected signals in densely vegetated areas, solving the signal annihilation and distance ambiguity problems caused by branch attenuation in traditional millimeter-wave radars in forest environments, and providing a distortion-free data foundation for subsequent waveform distortion elimination and three-dimensional reconstruction, enabling drones to maintain modeling integrity in areas with extremely high vegetation coverage.

[0049] These and other aspects of the present invention will become more readily apparent from the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1 A flowchart of a method for modeling a three-dimensional environment of a UAV based on millimeter-wave radar provided in an embodiment of the present invention;

[0052] Figure 2 A schematic structural diagram of a millimeter-wave radar-based UAV three-dimensional environment modeling system provided in an embodiment of the present invention;

[0053] Figure 3 A schematic diagram of the structure of a computing device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0054] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0055] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., and do not represent the order of precedence, nor do they limit "first" and "second" to be different types.

[0056] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0057] Figure 1 The present invention provides a flowchart of a method for modeling a three-dimensional environment of a UAV based on a millimeter-wave radar. Figure 1 As shown, the method includes:

[0058] In view of the serious attenuation of millimeter-wave radar signals and strong multipath reflection interference caused by dense vegetation in forest environments, and the inability of existing three-dimensional modeling technology to distinguish the material characteristics of vegetation, different structures such as trunks, branches, leaves, and vines are mixed into a single geometric body. At the same time, spatial positioning drift is caused by signal distortion, which ultimately leads to structural disorder, missing material information, blurred details and other defects in the reconstructed model. This makes it difficult for traditional drone surveys to obtain usable environmental models in areas with high vegetation coverage. To address this problem, the research and development ideas of the present invention are: first, to enhance the penetration of millimeter-wave signals through dynamic optimization of high-frequency circuits and multi-band fusion, to ensure the strength of deep vegetation reflection signals from the source; secondly, to introduce electromagnetic wave propagation The path inversion mechanism physically separates direct waves and multipath interference waves based on the wave equation, completely eliminating waveform distortion caused by multiple reflections of branches and leaves; then jointly analyzes the dielectric properties and spatial coordinates in the reflected signal, binds the physical properties of the material with the geometric structure, and converts biological characteristics such as vertical linear features of tree trunks and radial features of branches and leaves into topological connection relationships through a preset plant morphology rule library; finally, the orientation of the structural unit is corrected according to the direction of the coordinate vector in the spatial distribution information, and the three-dimensional model is combined according to the plant's phototropism and anti-gravity growth law to achieve cross-dimensional modeling from the signal layer to the physical layer, and can still output a forest space model with millimeter-level geometric accuracy and material semantic information in an environment shielded by branches and leaves. Based on this, the present invention provides a three-dimensional environment modeling method for unmanned aerial vehicles based on millimeter-wave radar, such as Figure 1 ,include:

[0059] Step 101: Utilize the millimeter wave radar pre-deployed in the UAV to transmit the optimized millimeter wave signal to a preset vegetation area to obtain a reflected signal from the vegetation area.

[0060] In this step, the optimized millimeter wave signal refers to the millimeter wave signal that adjusts the transmission efficiency through impedance matching, enhances the penetration through multi-band fusion, and improves the resolution through pulse compression, which is used to overcome the attenuation of vegetation signals; the vegetation area reflection signal refers to the echo signal formed after the optimized millimeter wave signal is reflected by the vegetation, which contains material characteristics and spatial distribution information.

[0061] In an embodiment of the present invention, first, the impedance matching parameters of the millimeter-wave radar transmitting circuit carried by the UAV are adjusted to optimize the signal transmission efficiency. Secondly, the multi-band signal source is driven to generate a combined frequency signal. Then, the combined frequency signal is waveform-shaped and pulse-width-compressed to generate an optimized millimeter-wave signal. Finally, the optimized millimeter-wave signal is directionally radiated to the vegetation area to receive the reflected echo, and the vegetation area reflection signal is generated after low-noise amplification.

[0062] Step 102: Reverse the signal propagation path of the vegetation area reflection signal to eliminate the waveform distortion feature in the vegetation area reflection signal, and generate a distortion-eliminated reflection signal.

[0063] In this step, signal propagation path inversion refers to the operation of tracing back the direction of electromagnetic waves based on phase changes, separating direct waves and reflected waves to eliminate multipath interference; waveform distortion characteristics refer to the signal phase aliasing and energy diffusion caused by multiple reflections of vegetation; the reflected signal after distortion is eliminated refers to the pure signal that retains the reflection characteristics of the real target after removing the interference beam.

[0064] In an embodiment of the present invention, the phase time series of the reflected signal in the vegetation area is first extracted to calculate the phase change rate, and then the propagation direction of the electromagnetic wave is inverted based on the phase change information to generate a direct wave path. Then, the wavefront energy distribution is reconstructed according to the direct wave path to separate the direct wave component and the reflected wave component. Finally, the multipath interference beam in the reflected wave component is removed to generate the reflected signal after distortion.

[0065] Step 103: Fusing the distortion-eliminated reflection signal with the vegetation material features in the vegetation area reflection signal to generate the three-dimensional structural features of the preset vegetation area.

[0066] In this step, the vegetation material feature refers to the dielectric constant parameter extracted from the reflected signal, which is used to identify the material type of tree trunks, branches and leaves; the fusion operation refers to the process of binding the position coordinates with the material structure rules to generate a topological connection relationship; the three-dimensional structure feature refers to the plant branch geometric model unit generated by fusing the position coordinates with the material rules.

[0067] In an embodiment of the present invention, the distance and azimuth information in the reflected signal after distortion is eliminated is first analyzed to generate a target position coordinate set, and then the dielectric constant parameters in the reflected signal of the vegetation area are extracted to match the preset material threshold interval to determine the material type identification, and then the plant structure rule library is queried to obtain the connection rules, and finally the material type identification and the connection rules are combined to generate three-dimensional structural features.

[0068] Step 104: spatially reconstructing the three-dimensional structural features based on the spatial distribution information in the reflection signal of the vegetation area to generate a three-dimensional spatial model of the preset vegetation area.

[0069] In this step, the spatial distribution information refers to the spatial position data set formed by the three-dimensional coordinate point set in the reflection signal; the spatial reconstruction operation refers to the process of correcting the orientation of the geometric units according to the plant growth law and combining them into a complete model.

[0070] In an embodiment of the present invention, the three-dimensional coordinate point set in the spatial distribution information is first parsed to construct a topological map containing position nodes. Secondly, the geometric units in the three-dimensional structural features are mapped to the position nodes of the topological map to generate a mapping relationship. Subsequently, the spatial orientation of the geometric units is corrected according to the mapping relationship to make it conform to the plant's phototropism and anti-gravity characteristics. Finally, all the corrected units are combined to generate a three-dimensional spatial model of the forest environment.

[0071] For example, the millimeter-wave radar transmitting circuit first adjusts the impedance network parameters to optimize signal transmission. Low-frequency penetration and high-frequency resolution signals are combined to generate an optimized millimeter-wave signal for directional radiation to the oak forest area. Echoes reflected from branches and leaves are received and amplified to generate a reflection signal from the vegetation area. Next, the phase sequence of the reflected signal is extracted to invert the electromagnetic wave propagation direction. The wavefront energy is reconstructed to separate the direct wave from the trunk and the reflected wave from the branches and leaves. Secondary reflection interference is removed to generate a distortion-free reflection signal. The distance and azimuth of the distortion-free signal are then analyzed to generate a set of trunk coordinates. The material type is determined by matching the oak wood dielectric constant threshold. A rule library is queried to obtain vertical linear connection rules and coordinates are bound to generate a three-dimensional trunk structure. Finally, the spatially distributed coordinates are analyzed to construct a topological map. Trunk geometric units are mapped to height sequence position nodes, corrected to a vertical orientation, and combined into an oak tree model. Simultaneously, the radial connections of the branch and leaf units are processed to generate a complete three-dimensional spatial model of the forest.

[0072] The embodiments of the present invention enhance vegetation penetration capability by dynamically optimizing millimeter-wave signals, eliminate multipath interference through path inversion, generate a biological characteristic topological model by combining material features with geometric structure binding, and then reconstruct the three-dimensional space according to the laws of phototropism and anti-gravity. This significantly solves the complex problems of signal attenuation, material confusion, and model distortion in forest environments, and achieves semantic three-dimensional reconstruction with centimeter-level accuracy in dense vegetation areas.

[0073] To address the issue of severe millimeter-wave signal attenuation caused by foliage obstruction in forest environments, this step improves signal transmission efficiency by dynamically optimizing impedance matching, fuses multi-band signals to enhance penetration, compresses pulse width to improve resolution, and combines directional radiation with echo amplification technology to generate high-quality vegetation reflection signals. The present invention provides a specific embodiment, step 101, utilizing a millimeter-wave radar pre-deployed in a drone to transmit an optimized millimeter-wave signal into a preset vegetation area to obtain a vegetation area reflection signal, specifically comprising the following steps:

[0074] Step 111: Adjust the impedance matching network parameters in the pre-deployed millimeter-wave radar to change the signal transmission characteristics of the transmitting circuit in the millimeter-wave radar, and generate optimized impedance parameters.

[0075] In this step, the impedance matching network parameters refer to the control variables for adjusting the capacitance and inductance values ​​of the transmitting circuit, which are used to maximize the transmission of RF signal energy; the signal transmission characteristics refer to the energy loss rate and phase stability of electromagnetic waves in the transmission line, reflecting the quality of the signal; the optimized impedance parameters refer to the capacitance and inductance combination value that minimizes the signal voltage standing wave ratio, which is used to eliminate signal reflection loss.

[0076] In an embodiment of the present invention, the preset impedance matching network parameters in the millimeter-wave radar transmitting circuit are first adjusted, and the RF channel impedance value is optimized by changing the capacitance and inductance values. Secondly, the optimal impedance matching point is calculated based on the change in signal transmission efficiency, and then the optimized impedance parameters are generated for subsequent signal synthesis.

[0077] Step 112: driving preset millimeter-wave signal sources of multiple frequency bands based on the optimized impedance parameters to generate a combined frequency signal.

[0078] In this step, the frequency band millimeter wave signal source refers to the hardware components of the preset low-frequency penetrating signal source and high-frequency resolution signal source; the combined frequency signal refers to the wide-band signal formed by the superposition of the low-frequency carrier and the high-frequency modulation wave, which has both penetration and resolution.

[0079] In an embodiment of the present invention, the optimized impedance parameters are first loaded into the signal source driving circuit, and then the preset low-frequency penetrating signal source and high-frequency resolution signal source are activated simultaneously, and then the two signals are carrier-superimposed to generate a combined frequency signal.

[0080] Step 113: Perform waveform shaping processing on the combined frequency signal to compress the pulse width of the combined frequency signal to generate an optimized millimeter wave signal.

[0081] In this step, waveform shaping refers to the operation of reducing the signal oscillation period through time-domain pulse compression technology; pulse width refers to the duration from the rising edge to the falling edge of the millimeter wave signal, which determines the distance resolution.

[0082] In an embodiment of the present invention, time domain waveform compression processing is first performed on the combined frequency signal, and then the signal oscillation time is reduced by rising edge acceleration and falling edge truncation operations, and then an optimized millimeter wave signal with pulse width compression is generated.

[0083] Step 114: Directly radiate the optimized millimeter wave signal to a preset vegetation area to trigger a vegetation reflection echo, amplify the vegetation reflection echo signal, and generate a vegetation area reflection signal.

[0084] In this step, the directional radiation operation refers to the action of focusing the signal energy to the preset vegetation area using beamforming technology; the vegetation reflection echo refers to the attenuated version of the signal formed by scattering on the plant surface, which contains material and structural information.

[0085] In an embodiment of the present invention, the optimized millimeter wave signal is first directionally focused to the vegetation area through the millimeter wave radar transmitting antenna, and then the scattered echo reflected by the leaves and trunks is received, and then the original echo signal is amplified by multiple levels of gain to generate the vegetation area reflection signal.

[0086] The embodiments of the present invention improve signal transmission efficiency by dynamically optimizing impedance matching, fuse dual-band signals to enhance vegetation penetration, compress pulse width to improve distance resolution, and combine directional radiation with echo amplification technology to significantly solve core problems such as signal attenuation, insufficient signal-to-noise ratio, and distance ambiguity in forest environments, providing a lossless data foundation for high-precision three-dimensional reconstruction.

[0087] To address the waveform aliasing distortion caused by multiple reflections from vegetation, this step captures the dynamic characteristics of the wavefront through phase change rate analysis, accurately locates the direct wave path using phase difference inversion, reconstructs the wavefront energy to separate the interference component, and rigidly removes the interference beam based on energy intensity differences to generate a pure reflection signal. The present invention provides a specific embodiment, step 102, which reverses the signal propagation path of the vegetation area reflection signal, eliminates the waveform distortion characteristics in the vegetation area reflection signal, and generates a distortion-free reflection signal, specifically including the following steps:

[0088] Step 201: Calculate the phase change rate of the phase time series in the vegetation area reflection signal to generate signal phase change information.

[0089] In this step, the phase time series refers to the set of signal phase angle values ​​recorded by the millimeter-wave radar receiving unit in chronological order, reflecting the process of electromagnetic wave phase change over time; the phase change rate refers to the amount of phase angle change per unit time, which is calculated by dividing the phase difference between adjacent time points by the sampling interval; the signal phase change information refers to a data set containing the phase change rate and change trend, which is used to characterize the dynamic characteristics of wavefront propagation.

[0090] In an embodiment of the present invention, the phase time series of the reflected signal in the vegetation area is first obtained, and then the phase angle difference between adjacent time sampling points is calculated. Then, the phase angle difference is divided by the time interval to generate the phase change rate, and finally the signal phase change information is output for path inversion.

[0091] Step 202: Calculate the phase difference between adjacent receiving units in the millimeter-wave radar based on the signal phase change information to invert the electromagnetic wave propagation direction of the optimized millimeter-wave signal and generate a direct wave propagation path.

[0092] In this step, the phase difference value refers to the phase angle difference between adjacent receiving units at the same time, reflecting the difference in the path length of electromagnetic waves reaching different antennas; the inversion operation refers to the physical derivation process of calculating the incident direction of the electromagnetic wave based on the geometric relationship between the phase difference and the antenna spacing; the direct wave propagation path refers to the spatial trajectory of the electromagnetic wave directly reflected from the target object to the radar, which is determined by the incident angle and distance.

[0093] In an embodiment of the present invention, the phase difference between adjacent receiving antennas of the millimeter-wave radar is first calculated based on the signal phase change information. Then, the phase difference is divided by the product of the antenna spacing and the wavelength to obtain the direction cosine value. Then, the incident angle of the electromagnetic wave is inverted to generate the direct wave propagation path.

[0094] Step 203: reconstructing the preset wavefront energy distribution according to the direct wave propagation path to generate a direct wave component and a reflected wave component.

[0095] In this step, the preset wavefront energy distribution refers to the wavefront surface energy density gradient model preset according to the wave equation theory; the reconstruction operation refers to the process of updating the wavefront curvature and energy distribution based on the direct wave path; the direct wave component refers to the real target reflection signal component with continuous energy gradient and consistent propagation direction; the reflected wave component refers to the discontinuous energy signal component containing multipath interference.

[0096] In an embodiment of the present invention, the wavefront curvature radius is first calculated based on the direct wave propagation path, and then the wavefront energy gradient distribution is reconstructed. Then, the energy continuous area and the discontinuous area are separated to generate the direct wave component and the reflected wave component.

[0097] Step 204: Compare the energy intensity difference between the direct wave component and the reflected wave component, remove the interference beam in the reflected wave component, and generate a distortion-eliminated reflected signal.

[0098] In this step, the energy intensity difference refers to the proportional relationship between the average energy values ​​of the direct wave and the reflected wave, which is used to identify interference; the interference beam refers to a false reflection signal in which the energy intensity of the reflected wave component is abnormally higher than that of the direct wave.

[0099] In an embodiment of the present invention, the average energy intensity ratio of the direct wave component and the reflected wave component is first compared, and then the beam in the reflected wave component whose energy intensity exceeds the set threshold of the direct wave component is marked as an interference beam, and then the interference beam is removed to generate a reflected signal after distortion is eliminated.

[0100] The embodiments of the present invention capture the dynamic characteristics of the wavefront through phase change rate analysis, use phase difference inversion to accurately locate the direct wave path, combine wavefront energy reconstruction to separate real reflections from multipath interference, and then rigidly eliminate interference beams based on energy intensity differences, completely solving the signal aliasing problem caused by multiple reflections of branches and leaves in forest environments, and providing a distortion-free data source for material feature extraction.

[0101] To address the difficulty in separating the direct and reflected waves due to wavefront distortion, this step quantifies the wavefront curvature using curvature parameters, maps the energy gradient into a three-dimensional spatial field, and identifies energy mutation boundaries based on physical thresholds to achieve a hard separation of the direct and reflected waves. The present invention provides a specific embodiment, step 203, which reconstructs the preset wavefront energy distribution based on the direct wave propagation path to generate the direct and reflected wave components. This specifically includes the following steps:

[0102] Step 231: Calculate the preset radius of the wavefront curvature based on the direct wave propagation path to generate wavefront curvature parameters.

[0103] In this step, the preset wavefront curvature refers to the reference value of the wavefront surface curvature set according to the wave equation theory, which is used to measure the wavefront energy diffusion characteristics; the wavefront curvature parameter refers to the curvature radius value calculated by the direct wave path length and direction change, which reflects the actual curvature state of the wavefront.

[0104] In an embodiment of the present invention, the wavefront curvature radius is first calculated based on the direct wave propagation path, and then the direct wave path length is multiplied by the wave velocity and then divided by the change in the path direction to obtain the curvature radius value. The calculated value is then scaled according to a preset wavefront curvature model, and finally the wavefront curvature parameters are output for energy field reconstruction.

[0105] Step 232: Mapping the spatial energy gradient of the wavefront curvature parameter to generate a reconstructed wavefront energy field.

[0106] In this step, the spatial energy gradient refers to the change in wavefront energy density within a unit distance, which is calculated by the inverse of the square of the curvature radius; mapping processing refers to the operation of assigning energy gradient values ​​to a three-dimensional spatial coordinate grid to form an energy distribution map; reconstructing the wavefront energy field refers to a three-dimensional matrix containing the energy gradient values ​​of each point in space, which represents the wavefront energy distribution state.

[0107] In an embodiment of the present invention, the curvature radius value in the wavefront curvature parameter is first obtained, and then the inverse of the square of the curvature radius is calculated as the spatial energy gradient reference value, and then the reference value is mapped to the three-dimensional spatial coordinate grid to generate the energy gradient distribution, and finally the reconstructed wavefront energy field is output for boundary scanning.

[0108] Step 233: Scan the preset energy continuity boundary conditions in the reconstructed wavefront energy field to generate energy mutation positions.

[0109] In this step, the preset energy continuity boundary condition refers to the threshold rule for determining whether the energy changes suddenly, including the maximum allowable change rate and jump frequency; the energy mutation position refers to the set of spatial coordinate points where the energy gradient value exceeds the continuity threshold, identifying the wavefront distortion area.

[0110] In an embodiment of the present invention, the energy value is first scanned in a preset direction in the reconstructed wavefront energy field, and then the energy change rate of adjacent grid cells is detected. Then, the area where the change rate exceeds the preset continuity boundary condition is marked as an energy mutation position, and finally the mutation position set is output for region segmentation.

[0111] Step 234: Divide the reconstructed wavefront energy field into continuous energy regions and discontinuous energy regions based on the energy mutation position, and generate direct wave components and reflected wave components.

[0112] In this step, the continuous energy region refers to the spatial region where the energy gradient changes smoothly, corresponding to the direct wave propagation path; the discontinuous energy region refers to the spatial region where the energy gradient jumps sharply, corresponding to multipath reflection interference.

[0113] In an embodiment of the present invention, the energy mutation position is first used as the boundary dividing line, and then the area where the energy change rate is lower than the threshold is divided into a continuous energy area, and then the energy jump area is divided into a discontinuous energy area. Finally, the continuous area is mapped as a direct wave component and the discontinuous area is mapped as a reflected wave component.

[0114] The embodiment of the present invention accurately quantifies the wavefront bending characteristics through curvature parameters, maps the energy gradient into a three-dimensional spatial field, identifies the energy mutation boundary based on the physical threshold, realizes the rigid separation of the direct wave and the reflected wave, and completely solves the wavefront distortion problem caused by multiple reflections of branches and leaves in the forest environment, providing physical layer protection for signal purification.

[0115] To address the disconnect between geometric structure and physical properties in forest scenes, this step generates a coordinate set by accurately analyzing the signal's spatiotemporal information, combines it with a material rule library to bind physical properties to spatial locations, constructs a topological connection network based on plant growth patterns, and outputs a semantic three-dimensional structure through geometric unit synthesis. The present invention provides a specific embodiment, step 103, which fuses the dedistorted reflection signal with the vegetation material features in the vegetation area reflection signal to generate the preset three-dimensional structural features of the vegetation area, specifically including the following steps:

[0116] Step 301: Analyze the distance and azimuth information in the distortion-eliminated reflected signal to generate a target position coordinate set.

[0117] In this step, the distance and azimuth information refers to the delay difference and phase difference data sets contained in the reflected signal. The delay difference reflects the target distance, and the phase difference reflects the azimuth angle. The analytical operation refers to the data processing process of calculating the straight-line distance from the delay difference and the azimuth angle from the phase difference. The target position coordinate set refers to the three-dimensional space point set generated by the analytical operation, which contains the longitude, latitude and altitude values ​​of the target.

[0118] In an embodiment of the present invention, the time delay difference and phase difference data are first extracted from the reflected signal after distortion is eliminated. Then, the time delay difference is multiplied by the speed of light and divided by two to calculate the straight-line distance. Then, the phase difference is divided by the wavelength and then multiplied by the antenna spacing to calculate the azimuth angle. Finally, the distance and azimuth values ​​are combined to generate a target position coordinate set.

[0119] Step 302: Match the vegetation material features in the vegetation area reflection signal with a preset material structure rule library to generate material structure attributes.

[0120] In this step, the preset material structure rule library refers to the database that stores dielectric threshold intervals and plant connection rules, including vertical linear rules for tree trunks and radial rules for branches and leaves; the matching operation refers to the query process of comparing the material dielectric constant with the rule library threshold interval to determine the material type; the material structure attribute refers to the structured data packet that binds the material type identifier and the connection rule.

[0121] In an embodiment of the present invention, the real part value of the dielectric constant in the vegetation material characteristics is first obtained, and then the dielectric threshold interval table in the preset material structure rule library is queried to determine the material type identifier, and then the plant structure rule entry is retrieved according to the material type identifier to obtain the connection rule, and finally the material type identifier and the connection rule are bound to generate the material structure attribute.

[0122] Step 303: constructing a spatial topological connection relationship of each position coordinate in the target position coordinate set based on the target position coordinate set and the material structure attribute.

[0123] In this step, the construction operation refers to the topological network generation action of connecting spatial coordinate points according to the material structure attribute rules; the spatial topological connection relationship refers to the set of connection paths between coordinate points, such as the vertical links of trunk nodes or the radial links of branch nodes.

[0124] In an embodiment of the present invention, the three-dimensional coordinate points in the target position coordinate set are first read, and then the topology construction strategy is determined according to the connection rule type in the material structure attribute, and then the trunk coordinate points are connected according to the vertical linear rule or the branch and leaf coordinate points are connected according to the radial rule, and finally a spatial topological connection relationship network is generated.

[0125] Step 304: traverse the connection nodes in the spatial topological connection relationship to synthesize three-dimensional geometric structure units and generate three-dimensional structural features.

[0126] In this step, the synthesis operation refers to the action of assembling geometric units into a complete structure in a topological path sequence; the three-dimensional geometric structure unit refers to the standardized geometric body generated according to the plant parts, such as the cylindrical unit representing the trunk.

[0127] In an embodiment of the present invention, each connection node in the spatial topological connection relationship is first traversed, and then the cylinder generation algorithm or the surface patch generation algorithm is called according to the node type, and then the geometric units are combined in the order of the connection path, and finally the three-dimensional structural features including the branch structure are output.

[0128] The embodiment of the present invention generates a coordinate set by accurately analyzing the signal's spatiotemporal information, combines it with a material rule library to achieve a rigid binding of physical properties and spatial positions, constructs a topological connection network according to the laws of plant growth, and then synthesizes and outputs a biologically characterized three-dimensional structure through geometric unit synthesis, completely solving the problems of material structure confusion and geometric distortion in forest scenes and achieving semantic modeling with millimeter-level precision.

[0129] To address the disconnect between material properties and spatial structure rules, this step achieves quantitative material characterization through dielectric parameter normalization, accurately distinguishes vegetation types using hard threshold matching, and, in conjunction with a plant rule library query, obtains biologically characterized connection rules to generate a data package that integrates physical properties and spatial structure. The present invention provides a specific embodiment, step 302, in which the vegetation material characteristics in the reflected signal of the vegetation area are matched with a preset material structure rule library to generate material structure properties. This specifically includes the following steps:

[0130] Step 321: extracting material electrical property parameters of vegetation material characteristics from the vegetation area reflection signal, converting the material electrical property parameters into standardized identifiers, and generating a material electrical identifier.

[0131] In this step, the material electrical property parameter refers to the real part of the complex dielectric constant extracted from the reflection signal, which reflects the electromagnetic wave reflection characteristics of the vegetation material; the conversion operation refers to the data processing process of mapping the original real part of the dielectric constant to the standard range through the normalization formula; the standardized identifier refers to the dielectric constant calibration value between zero and one after normalization processing; the material electrical identifier refers to the standardized identifier output value representing the electromagnetic characteristics of the material.

[0132] In an embodiment of the present invention, the real part of the dielectric constant of the vegetation material characteristics in the reflection signal of the vegetation area is first extracted as the material electrical property parameter, and then the real part of the dielectric constant is subtracted from the preset minimum threshold and divided by the difference between the maximum threshold and the minimum threshold to achieve normalization processing, and then the normalization result is quantized into a calibration value in the range of zero to one to generate a material electrical identification.

[0133] Step 322: Match the material electrical identifier with a preset material threshold range to determine the material type identifier of the material electrical identifier.

[0134] In this step, the preset material threshold range refers to the dielectric constant calibration value classification standard divided by forest vegetation type, including the lower limit of the trunk threshold and the upper limit of the branch threshold; the matching operation refers to the logical judgment process of comparing the material electrical identification with the threshold range; the material type identification refers to the material classification label output by the matching result, including types such as trunks, branches and leaves.

[0135] In an embodiment of the present invention, the calibration value of the material electrical identification is first read, and then the calibration value is compared with the preset material threshold range. Then, if the calibration value is greater than or equal to 0.6, it is determined to be the trunk material type identification; if the calibration value is less than or equal to 0.3, it is determined to be the branch and leaf material type identification, and finally the determined material type identification is output.

[0136] Step 323: querying a preset plant structure rule library based on the material type identifier to obtain a structure connection rule corresponding to the material type identifier.

[0137] In this step, the preset plant structure rule library refers to a database that stores spatial connection rules corresponding to different material types; the query operation refers to the data call process of retrieving rule library entries based on material type identifiers; and the structural connection rules refer to text instructions that describe the spatial connection method of geometric units, such as vertical linear connection.

[0138] In an embodiment of the present invention, first, a preset plant structure rule library is accessed according to the material type identifier. Secondly, if the material type identifier is a trunk, the vertical linear connection rule entry is queried; if it is a branch or leaf, the radial connection rule entry is queried. Then, the rule description text is extracted to generate the structural connection rule.

[0139] Step 324: Combine the material type identifier with the structural connection rule to generate material structural attributes.

[0140] In this step, the combination operation refers to a data encapsulation action of binding the material type identifier and the structural connection rule into a key-value pair.

[0141] In an embodiment of the present invention, a data structure including a material type identifier and a structure connection rule is first created, and then the material type identifier is used as a key name and the structure connection rule is used as a key value for pairing and binding, and then the material structure attributes are generated by encapsulating the data into a structured data packet.

[0142] The embodiment of the present invention realizes the quantitative characterization of material properties through dielectric parameter normalization, uses hard threshold matching to accurately distinguish vegetation types, combines plant rule library queries to obtain biological characteristic connection rules, and finally generates a data package that integrates physical properties and spatial structure, completely solving the problem of separation between material properties and geometric structure in forest scenes, and providing semantic input for three-dimensional modeling.

[0143] To address the misalignment between the geometric model and the actual spatial distribution, this step constructs a quantitative vegetation distribution pattern using a topological map, accurately positions geometric units based on node mapping, corrects component orientation using gravity and illumination vectors, and generates an ecologically compliant three-dimensional space based on the combined model of biological growth order. The present invention provides a specific embodiment, step 104, which spatially reconstructs the three-dimensional structural features based on the spatial distribution information in the reflected signal of the vegetation area to generate a three-dimensional spatial model of the preset vegetation area. Specifically, the steps include:

[0144] Step 401: Analyze the three-dimensional coordinate point set of the spatial distribution information in the reflection signal of the vegetation area to generate a spatial position topology map.

[0145] In this step, the three-dimensional coordinate point set refers to the set of longitude, latitude and altitude coordinates of all target objects in the spatial distribution information, which constitutes the basic spatial position data; the parsing operation refers to the process of establishing the connection relationship between coordinate points according to the growth law of plants, including the application of vertical connection and radial connection rules; the spatial position topology map refers to the network structure in which the position nodes carry coordinate information and the connection relationship describes the growth path, reflecting the spatial distribution form of vegetation.

[0146] In an embodiment of the present invention, the three-dimensional coordinate point set of the spatial distribution information in the reflection signal of the vegetation area is first read, and then a vertical connection relationship is established according to the height difference threshold to simulate the growth direction of the tree trunk, and then a radial connection is established according to the horizontal distance threshold to simulate the distribution of branches and leaves, and finally a spatial position topology map containing position nodes and connection relationships is generated.

[0147] Step 402: Map the three-dimensional geometric structure units in the three-dimensional structure feature to the position nodes in the spatial position topology graph to generate a node mapping relationship.

[0148] In this step, the mapping operation refers to the data association action of binding the geometric structure unit and the topological map location node by type; the node mapping relationship refers to the corresponding association table of the geometric unit and the location node, which records the matching entries of the unit identifier and the node coordinates.

[0149] In an embodiment of the present invention, the type identifier of the geometric structure unit in the three-dimensional structural feature is first identified, and then the position node matching the unit type is searched in the spatial position topology map, and then the trunk component unit is mapped to the height sequence node, and the branch and leaf component unit is mapped to the terminal branch node, and finally a node mapping relationship table is generated.

[0150] Step 403: adjusting the spatial orientation of the three-dimensional geometric structure unit according to the node mapping relationship to generate an orientation-corrected structure unit.

[0151] In this step, spatial orientation refers to the direction vector of the geometric structure unit in three-dimensional space, which is determined by the angle between the axis and the direction of gravity and the direction of light; the adjustment operation refers to the physical correction process of rotating the geometric unit so that its axis is aligned with the composite vector of gravity and light; the orientation correction structure unit refers to a standardized geometric component whose spatial orientation conforms to the growth characteristics of plants, such as a vertical trunk unit.

[0152] In an embodiment of the present invention, the spatial coordinate vector of the position node is first obtained according to the node mapping relationship, and then the synthetic reference vector of the gravity direction and the illumination direction is calculated. Then, the geometric structure unit is rotated so that its axis is aligned with the reference vector, and finally the orientation correction structure unit is output.

[0153] Step 404: combine all the orientation correction structural units to generate a three-dimensional spatial model of the preset vegetation area.

[0154] In this step, the combination operation refers to the spatial splicing action of assembling the correction units in the order of topological connection, following the rule of trunk first and branches second.

[0155] In an embodiment of the present invention, all orientation correction structural units are first traversed in the connection order of the spatial position topology diagram, then the trunk components are spliced ​​end to end along the vertical path, and then the branch and leaf components are radially connected to the corresponding trunk nodes, and finally combined into a complete three-dimensional spatial model of the forest environment.

[0156] The embodiment of the present invention constructs a quantitative vegetation spatial distribution pattern through a topological map, realizes the precise positioning of geometric units based on node mapping, corrects the orientation of components by combining gravity and light synthesis vectors, and finally combines the models according to the order of biological growth, completely solving the problems of geometric structure dislocation and direction misalignment in forest scenes, and outputting a three-dimensional spatial model that conforms to ecological laws.

[0157] Figure 2 The present invention provides a schematic diagram of the structure of a UAV three-dimensional environment modeling system based on millimeter wave radar, as shown in FIG. Figure 2 As shown, the system includes:

[0158] The transmitting module 21 is used to transmit the optimized millimeter wave signal to the preset vegetation area using the millimeter wave radar pre-deployed in the UAV to obtain the reflected signal of the vegetation area;

[0159] an elimination module 22 for inversely calculating a signal propagation path of the vegetation area reflection signal, eliminating waveform distortion features in the vegetation area reflection signal, and generating a distortion-eliminated reflection signal;

[0160] A fusion module 23 is configured to fuse the distortion-eliminated reflection signal with vegetation material features in the vegetation area reflection signal to generate a three-dimensional structural feature of the preset vegetation area;

[0161] The reconstruction module 24 is configured to spatially reconstruct the three-dimensional structural features based on the spatial distribution information in the reflection signal of the vegetation area, so as to generate a three-dimensional spatial model of the preset vegetation area.

[0162] Figure 2 The millimeter wave radar-based UAV three-dimensional environment modeling system can perform Figure 1 The implementation principles and technical effects of the millimeter-wave radar-based 3D environment modeling method for UAVs described in the illustrated embodiment will not be elaborated upon. The specific manner in which each module and unit performs operations in the millimeter-wave radar-based 3D environment modeling system for UAVs in the aforementioned embodiment has been described in detail in the relevant embodiments of the method and will not be elaborated upon here.

[0163] In one possible design, Figure 2 The millimeter wave radar-based UAV three-dimensional environment modeling system of the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0164] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0165] The processing component 32 is used to: use the millimeter wave radar pre-deployed in the drone to transmit the optimized millimeter wave signal to the preset vegetation area to obtain the vegetation area reflection signal; reverse the signal propagation path of the vegetation area reflection signal, eliminate the waveform distortion characteristics in the vegetation area reflection signal, and generate the distortion-eliminated reflection signal; fuse the distortion-eliminated reflection signal with the vegetation material characteristics in the vegetation area reflection signal to generate the three-dimensional structural characteristics of the preset vegetation area; spatially reconstruct the three-dimensional structural characteristics based on the spatial distribution information in the vegetation area reflection signal to generate a three-dimensional spatial model of the preset vegetation area.

[0166] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above method.

[0167] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile memory device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0168] Of course, a computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0169] The input / output interface provides an interface between the processing component and the peripheral interface module, which can be an output device, an input device, etc.

[0170] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0171] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0172] The embodiment of the present invention further provides a computer storage medium storing a computer program, which can achieve the above-mentioned Figure 1 The embodiment shown is a method for modeling a three-dimensional environment of an unmanned aerial vehicle based on millimeter-wave radar.

[0173] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0174] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0175] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0176] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for modeling a three-dimensional environment of an unmanned aerial vehicle based on millimeter-wave radar, characterized in that: include: The millimeter-wave radar pre-deployed in the drone is used to transmit the optimized millimeter-wave signal into the preset vegetation area to obtain the reflected signal of the vegetation area; Reversely inferring a signal propagation path of the vegetation area reflection signal, eliminating waveform distortion features in the vegetation area reflection signal, and generating a distortion-eliminated reflection signal; fusing the distortion-eliminated reflection signal with vegetation material features in the vegetation area reflection signal to generate three-dimensional structural features of the preset vegetation area; The three-dimensional structural features are spatially reconstructed based on the spatial distribution information in the reflection signal of the vegetation area to generate a three-dimensional spatial model of the preset vegetation area.

2. The method according to claim 1, characterized in that The millimeter-wave radar pre-deployed in the drone transmits the optimized millimeter-wave signal to the preset vegetation area to obtain the reflected signal of the vegetation area, including: Adjusting the impedance matching network parameters in a pre-deployed millimeter-wave radar to change the signal transmission characteristics of the transmitting circuit in the millimeter-wave radar and generate optimized impedance parameters; Driving preset multiple frequency band millimeter wave signal sources based on the optimized impedance parameters to generate a combined frequency signal; performing waveform shaping processing on the combined frequency signal to compress the pulse width of the combined frequency signal to generate an optimized millimeter wave signal; The optimized millimeter wave signal is directionally radiated into a preset vegetation area to trigger a vegetation reflection echo, and the signal of the vegetation reflection echo is amplified to generate a vegetation area reflection signal.

3. The method according to claim 1, characterized in that Reversely inferring a signal propagation path of the vegetation area reflection signal, eliminating waveform distortion features in the vegetation area reflection signal, and generating a distortion-eliminated reflection signal, including: Calculating the phase change rate of the phase time series in the vegetation area reflection signal to generate signal phase change information; Calculating phase differences between adjacent receiving units in the millimeter-wave radar based on the signal phase change information to invert the electromagnetic wave propagation direction of the optimized millimeter-wave signal and generate a direct wave propagation path; Reconstructing a preset wavefront energy distribution according to the direct wave propagation path to generate a direct wave component and a reflected wave component; The energy intensity difference between the direct wave component and the reflected wave component is compared, and the interference beam in the reflected wave component is eliminated to generate a distortion-eliminated reflected signal.

4. The method according to claim 3, characterized in that Reconstructing a preset wavefront energy distribution according to the direct wave propagation path to generate a direct wave component and a reflected wave component, including: Calculating a preset radius of wavefront curvature based on the direct wave propagation path to generate a wavefront curvature parameter; Mapping the spatial energy gradient of the wavefront curvature parameter to generate a reconstructed wavefront energy field; Scanning a preset energy continuity boundary condition in the reconstructed wavefront energy field to generate an energy mutation location; The reconstructed wavefront energy field is divided into continuous energy regions and discontinuous energy regions based on the energy mutation position to generate a direct wave component and a reflected wave component.

5. The method according to claim 1, wherein The distortion-eliminator reflected signal is fused with the vegetation material feature in the vegetation area reflected signal to generate the three-dimensional structural feature of the preset vegetation area, including: Analyzing the distance and azimuth information in the distortion-eliminated reflected signal to generate a target position coordinate set; Matching vegetation material features in the reflection signal of the vegetation area with a preset material structure rule library to generate material structure attributes; Constructing a spatial topological connection relationship between each position coordinate in the target position coordinate set based on the target position coordinate set and the material structure attribute; The connection nodes in the spatial topological connection relationship are traversed to synthesize three-dimensional geometric structure units and generate three-dimensional structural features.

6. The method according to claim 5, characterized in that Matching the vegetation material features in the vegetation area reflection signal with a preset material structure rule library to generate material structure attributes, including: Extracting material electrical property parameters of vegetation material characteristics from the vegetation area reflection signal, converting the material electrical property parameters into a standardized identifier, and generating a material electrical identifier; Matching the material electrical identification with a preset material threshold interval to determine a material type identification of the material electrical identification; Based on the material type identifier, a query is performed in a preset plant structure rule library to obtain a structural connection rule corresponding to the material type identifier; The material type identifier is combined with the structural connection rule to generate material structural attributes.

7. The method according to claim 1, characterized in that The three-dimensional structural features are spatially reconstructed based on the spatial distribution information in the reflection signal of the vegetation area to generate a three-dimensional spatial model of the preset vegetation area, including: parsing a three-dimensional coordinate point set of spatial distribution information in the reflection signal of the vegetation area to generate a spatial position topology map; Mapping the three-dimensional geometric structure units in the three-dimensional structure features to the position nodes in the spatial position topology graph to generate a node mapping relationship; Adjusting the spatial orientation of the three-dimensional geometric structure unit according to the node mapping relationship to generate an orientation-corrected structure unit; All the orientation correction structural units are combined to generate a three-dimensional spatial model of the preset vegetation area.

8. A millimeter-wave radar-based three-dimensional environment modeling system for unmanned aerial vehicles, characterized in that: include: The transmitting module is used to transmit the optimized millimeter wave signal to the preset vegetation area using the millimeter wave radar pre-deployed in the UAV to obtain the reflected signal of the vegetation area; an elimination module, configured to reversely infer a signal propagation path of the vegetation area reflection signal, eliminate waveform distortion features in the vegetation area reflection signal, and generate a distortion-eliminated reflection signal; a fusion module, configured to fuse the distortion-eliminated reflection signal with vegetation material features in the vegetation area reflection signal to generate three-dimensional structural features of the preset vegetation area; A reconstruction module is used to spatially reconstruct the three-dimensional structural features based on the spatial distribution information in the reflection signal of the vegetation area to generate a three-dimensional spatial model of the preset vegetation area.

9. A computing device, characterized in that It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a three-dimensional environment modeling method for unmanned aerial vehicle based on millimeter wave radar as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that A computer program is stored, and when the computer program is executed by a computer, the method for three-dimensional environment modeling of a UAV based on millimeter wave radar as described in any one of claims 1 to 7 is implemented.

Images