A lunar rover wheel dust-raising field in-situ dynamic observation method

Abstract

The application discloses a lunar rover wheel dust field in-situ dynamic observation method and belongs to the technical field of lunar exploration. In order to solve the defects in the prior art that there is no in-situ observation method which can be applied to the lunar surface environment, has the multi-sensor cooperation capability, and can realize the real-time, dynamic and three-dimensional reconstruction and characteristic analysis of the lunar rover wheel dust field, the technical scheme provided by the application comprises the following steps: synchronously processing image data, laser point cloud data and particle concentration particle size data, three-dimensional reconstruction, fusion analysis and modeling, and outputting the dynamic parameters of the dust field. The processing of the image data comprises image enhancement, feature extraction, feature matching on the image sequence, and the generation of three-dimensional point cloud by using multi-view triangulation. The generated point cloud is subjected to statistical filtering, noise points are removed, and a continuous surface model is constructed. The method is suitable for intuitively observing the lunar rover wheel dust track in the special lunar environment.

Description

Technical Field

[0001] This relates to the field of lunar exploration technology, specifically to a method for in-situ observation of dust emitted by lunar rover wheels. Background Technology

[0002] With the continuous advancement of my country's deep space exploration missions, autonomous exploration and scientific investigation by lunar rovers in the lunar environment has become one of the key tasks of the lunar exploration project. However, the lunar surface is covered with a layer of loose lunar regolith composed of fine particles, of which lunar dust with a diameter of less than 1 mm accounts for a very high proportion. These particles have irregular structures, sharp edges, and strong adhesion and mobility. Since the Moon's gravity is only about 1 / 6 that of Earth, and there is a lack of atmospheric damping, the interaction between the lunar rover's wheels and the lunar regolith during its operation easily generates dust. This dust not only pollutes the rover itself (e.g., adhering to solar panels, cameras, or optical instruments), but can also obstruct the field of vision, interfere with the normal execution of scientific exploration missions, and even affect wheel grip and stability. Therefore, conducting in-situ observations and pattern analysis of dust generation during lunar rover operation is of great significance for ensuring mission safety and optimizing vehicle design.

[0003] Currently, some research teams have attempted to analyze lunar dust ejection behavior from the perspectives of experimental simulation and remote sensing monitoring. For example, the Apollo program indirectly recorded dust ejection phenomena through visual recordings by astronauts and external cameras on the mission module. However, these records have limited spatial resolution and lack a systematic characterization of particle properties and spatial distribution. In recent years, some studies have also attempted to establish experimental platforms in a vacuum, low-gravity environment on Earth to simulate the dust ejection behavior of lunar regolith under different disturbances. However, these methods cannot fully reproduce the complex plasma and electric field conditions in the real lunar environment and their impact on the suspension and migration of lunar dust.

[0004] Regarding data acquisition methods, some studies have attempted to use lidar scanning or high-speed imaging techniques to observe the movement of dust particles. However, due to the lack of a systematically integrated multimodal sensing and efficient data processing mechanism, they often only capture local or static information, making it difficult to achieve high-precision, real-time, and three-dimensional reconstruction and analysis of the dynamic evolution of dust fields. In addition, traditional particle counters or spectral sensors are often used for dust monitoring inside space stations. In open spaces (such as the lunar surface), they face problems such as poor adaptability to particle size range, insufficient space coverage, and difficulties in data fusion, lacking an integrated observation framework for multi-dimensional and multi-scale particle scenarios.

[0005] In summary, the existing technology lacks an in-situ observation method that is applicable to the lunar surface environment, has multi-sensor collaboration capabilities, and can realize real-time, dynamic, three-dimensional reconstruction and characteristic analysis of lunar rover wheel dust field. Summary of the Invention

[0006] To address the shortcomings of existing technologies, such as the lack of an in-situ observation method suitable for the lunar surface environment, possessing multi-sensor collaboration capabilities, and capable of real-time, dynamic, three-dimensional reconstruction and characteristic analysis of lunar rover wheel dust fields, the technical solution provided by this invention is as follows: A method for in-situ dynamic observation of dust fields from lunar rover wheels includes: The steps involve simultaneously processing, 3D reconstructing, fusion analysis, and modeling image data, laser point cloud data, and particle concentration and size data to output dynamic parameters of the dust field.

[0007] Furthermore, a preferred embodiment is that the image data processing includes image enhancement, feature extraction, feature matching of the image sequence, and generation of a three-dimensional point cloud using multi-view triangulation.

[0008] Furthermore, a preferred embodiment involves performing statistical filtering on the generated point cloud to remove noise points and construct a continuous surface model.

[0009] Furthermore, a preferred embodiment is that the fusion analysis includes aligning the image point cloud and the laser point cloud in space, and unifying the estimation of particle size and concentration features.

[0010] Furthermore, in a preferred embodiment, the processing of particle size and concentration characteristics includes generating a concentration field and a particle size field using a spatial interpolation method, and extracting the concentration gradient and regional features.

[0011] A dynamic in-situ observation device for lunar rover wheel dust field is also provided, including: This module performs simultaneous processing, 3D reconstruction, fusion analysis, and modeling of image data, laser point cloud data, and particle concentration and size data, and outputs dynamic parameters of the dust field.

[0012] A lunar rover wheel dust field in-situ dynamic observation system is also provided to implement the method, including: A high-speed camera mounted on the dust baffle is used to capture dynamic images of dust. A particle counting sensor installed on the side of the vehicle body is used to detect the concentration and size of dust particles; A lidar unit mounted on the vehicle body is used to acquire spatial distribution information of dust. A global camera mounted high on the vehicle body is used to capture overall images of the dust field.

[0013] A computer storage medium is also provided for storing a computer program, which, when read by the computer, executes the method.

[0014] A computer is also provided, including a processor and a storage medium, wherein the computer executes the method when the processor reads a computer program stored in the storage medium.

[0015] A computer program product is also provided, which, when executed, implements the method described.

[0016] Compared with the prior art, the advantages of the technical solution provided by the present invention are as follows: This approach integrates multiple sensors, including lidar, high-speed cameras, spectrometers, and particle counters, onto the lunar rover to construct a multi-source collaborative observation system, enabling comprehensive data acquisition of dust fields generated by the rover's wheels. Compared to existing methods that use only a single sensor for dust detection, this approach can simultaneously acquire information on spatial distribution, particle size concentration, and motion trajectory, enhancing the overall perception of dust fields and providing a rich data foundation for subsequent analysis and modeling.

[0017] By employing a fusion observation method combining high-frame-rate high-speed cameras and lidar, along with computer vision and 3D reconstruction techniques, a dynamic 3D model of the dust field was constructed. This method can accurately capture the movement trajectory and diffusion range of lunar dust over time after disturbance. Compared with traditional research methods that rely on static images or 2D projection analysis, it has higher spatiotemporal resolution and stronger on-site reconstruction capabilities, and can intuitively reflect the dynamic characteristics of dust behavior in real-world environments.

[0018] By performing spatial interpolation and statistical analysis on dust spatial data measured by lidar and particle size and concentration data output by particle counters, continuous particle size and concentration field distribution maps are constructed, and the external forces acting on the particles can be inverted. Compared with existing methods that are limited to fixed-point sampling or average concentration statistics, this scheme achieves multi-scale correlation from spatial distribution to mechanical behavior, and can systematically reveal the dust particle migration mechanism and explain the physical causes of the high probability of fine particle suspension.

[0019] In terms of data fusion, this solution establishes a feature mapping and similarity association mechanism based on trajectory features, particle size features, concentration features, and motion parameters extracted from different sensors, and constructs a multi-input neural network fusion model to ultimately output complete comprehensive parameters of the dust field. This approach overcomes the bottlenecks of limited data fusion dimensions and low fusion accuracy in existing technologies, achieving effective fusion of high-precision, multi-scale, and multi-source data, which helps to significantly improve analytical accuracy and modeling and predictive capabilities.

[0020] In terms of displaying observation results, the dynamic dust emission process is visualized using 3D graphics, intuitively reflecting the diffusion trend of dust particles under vehicle disturbance. Compared with traditional tabular data or two-dimensional image presentation methods, this display method significantly improves the efficiency of interpreting observation results and the reference value for scientific research decision-making, providing a more intuitive and reliable basis for task scheduling and protection design.

[0021] It is suitable for visually observing the dust trails of lunar rover wheels in the special environment of the moon. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the observation system. Figure 2 for Figure 1 Axonometric drawing; Figure 3 This is a flowchart for the 3D reconstruction of a dust field.

[0023] Among them, 1 is a high-speed camera, 2 is a particle counting sensor, 3 is a lidar, 4 is a lunar rover dust shield, and 5 is a global camera. Detailed Implementation

[0024] To make the advantages and benefits of the technical solution provided by the present invention clearer, the technical solution provided by the present invention will now be described in further detail with reference to the accompanying drawings, specifically: Implementation Method 1: This implementation method provides an in-situ dynamic observation method for dust field generated by lunar rover wheels, including: The steps involve simultaneously processing, 3D reconstructing, fusion analysis, and modeling image data, laser point cloud data, and particle concentration and size data to output dynamic parameters of the dust field.

[0025] Image data processing includes image enhancement, feature extraction, and feature matching of image sequences, and the generation of three-dimensional point clouds using multi-view triangulation.

[0026] Statistical filtering is applied to the generated point cloud to remove noise points and construct a continuous surface model.

[0027] Fusion analysis includes spatially aligning image point clouds with laser point clouds and unifying particle size and concentration feature estimation.

[0028] The processing of particle size and concentration characteristics includes generating concentration and particle size fields using spatial interpolation methods, and extracting concentration gradients and regional features.

[0029] A lunar rover wheel dust field in-situ dynamic observation system is also provided to implement the method, including: A high-speed camera mounted on the dust baffle is used to capture dynamic images of dust. A particle counting sensor installed on the side of the vehicle body is used to detect the concentration and size of dust particles; A lidar unit mounted on the vehicle body is used to acquire spatial distribution information of dust. A global camera mounted high on the vehicle body is used to capture overall images of the dust field.

[0030] Implementation Method Two: This implementation method is a further detailed description of the technical solution provided in Implementation Method One, specifically: A method for in-situ dynamic observation of lunar rover wheel dust field is proposed. This method constructs a multi-sensor fusion observation system to address the wheel dust behavior under the low gravity, vacuum, and complex electromagnetic environment of the lunar surface. Through step-by-step data acquisition, preprocessing, 3D reconstruction, characteristic analysis, and data fusion modeling, dynamic, 3D, and high-precision monitoring and analysis of the dust field can be achieved.

[0031] First, the observation system was built. Multiple sensor components are installed on the lunar rover's main structure, including a lidar system (such as the Velodyne VLP-16), a high-speed camera, a spectrometer (such as the Ocean Optics Maya2000 Pro), and a particle counter (such as the TSI AeroTrak9306). The lidar is used to acquire information on the terrain and spatial distribution of dust in the area near the rover's wheels. The high-speed camera is used to continuously record the dynamic process of dust and the trajectory of particles. The particle counter is used to monitor the concentration and size distribution of lunar dust particles at specific spatial points. In addition, dust baffles and corresponding mounting holes are installed to stabilize the camera devices; a global camera is used to cover a wide area of ​​dust field to achieve overall environmental recording.

[0032] Secondly, data collection and transmission were carried out. The lunar rover is activated and slowly moves along a predetermined trajectory, while various sensors are simultaneously turned on. The lidar periodically emits laser pulses and receives echoes, forming three-dimensional point cloud data of the area around the rover; a high-speed camera records image sequences of particle changes over time at a high frame rate; and a particle counter collects and records the number and size distribution of lunar dust particles per unit volume. The collected raw data is transmitted in real time to the central processing system via a communication module for further processing.

[0033] Then, perform data preprocessing. The central processing system performs unified preprocessing operations on laser, image, and particle data, including outlier removal, noise filtering, coordinate correction, and temporal alignment. Specifically, image data undergoes edge enhancement and deblurring algorithms to improve particle boundary clarity; laser data is filtered and denoised based on statistical distances between laser points; and particle counter data is standardized for concentration and particle size. This step ensures the accuracy and consistency of multi-source data, laying the foundation for subsequent analysis.

[0034] Next, a three-dimensional reconstruction of the dust field was carried out. By combining point cloud data obtained from LiDAR with high-speed camera image data, a three-dimensional spatial model of the dust field is reconstructed using computer vision and 3D reconstruction algorithms. First, the high-speed camera module is precisely calibrated to determine its spatial position and orientation in the world coordinate system. Then, particle feature points are extracted from the image sequence, and matching algorithms (such as those based on scale-invariant or speed-robust features) are used to identify corresponding points of the same particle in multiple viewpoint images. Multi-view triangulation is then used to calculate the particle positions in three-dimensional space, forming a preliminary point cloud. Finally, statistical filtering is used to remove outliers, and surface reconstruction algorithms are employed to connect the particle cloud into a continuous surface, constructing a complete dust field morphology.

[0035] Subsequently, dust characteristics analysis was conducted. Based on particle size and concentration data output from a particle counter, and combined with corresponding measurement points in the laser point cloud, spatial interpolation is performed to generate continuous concentration and particle size fields. High-density dust areas are located using concentration gradient calculations. By analyzing the differences in particle size distribution at different heights and combining this with changes in particle velocity, the external forces acting on the particles, such as gravity, electric field force, thrust, and inter-particle collision forces, are deduced, allowing for the analysis of dynamic response differences among particles of different sizes. This process helps to understand why fine particles are more easily suspended and remain in the air for longer periods.

[0036] Next, data fusion and modeling are implemented. Key feature sets are selected from the features extracted from various sensors, such as particle velocity, direction, and morphology in camera images, particle size and concentration in laser point clouds, and quantity and size distribution statistics in particle counters. Cross-matching is performed using feature similarity calculation methods to establish mapping relationships. A filtering algorithm is employed to correct conflicting data; when the error exceeds a set threshold, data with higher weight is selected as the reference value. Finally, all features are input into a constructed multi-input neural network model to predict and dynamically model the dust field state, outputting fused multi-dimensional parameters such as concentration, particle size, and motion vectors.

[0037] Finally, the results can be visualized and displayed. The 3D model of the dust field, particle concentration distribution, velocity vectors, and other data are graphically rendered to construct a dynamic 3D visualization scene, allowing for interactive observation of the dust field's morphological evolution from multiple perspectives. This visualization result can intuitively support researchers' understanding of lunar dust disturbance behavior and provide a basis for decision-making in the subsequent design and optimization of lunar rover operation strategies.

[0038] Implementation Method 3: Combination Figure 1-3 This embodiment describes the technical solution provided above in further detail through specific examples. Specifically: This implementation aims to provide an in-situ dynamic observation method for lunar rover wheel dust fields based on multi-sensor fusion. This method enables real-time, dynamic, and high-precision observation of dust fields, providing crucial data support for lunar rover exploration and related research. The specific steps are as follows: Observation system setup: The lunar rover was equipped with various sensors, including a Velodyne VLP-16 lidar, a high-speed camera, an Ocean Optics Maya2000 Pro spectrometer, and a TSI AeroTrak 9306 particle counter. The lidar was used to scan the dust distribution and terrain information around the rover's wheels; the high-speed camera was used to capture the movement and morphological changes of the dust; and the particle counter was used to detect the concentration and size distribution of dust particles.

[0039] Data Acquisition and Transmission: The lunar rover is activated and driven along a predetermined route and speed, while its various sensors are engaged to collect data. The lidar emits laser pulses at regular intervals and receives reflected signals to acquire three-dimensional spatial distribution data of dust around the rover's wheels; a high-speed camera continuously captures dust images at a high frame rate; and a particle counter collects information on the concentration and size of dust particles. The acquired data is transmitted in real-time to the rover's central processing system via a data transmission module.

[0040] Data preprocessing: The central processing system preprocesses the received data, including data cleaning, filtering, and correction, to remove noise and abnormal data and improve the accuracy and reliability of the data.

[0041] 3D Reconstruction of Dust Fields: Based on 3D spatial distribution data collected by LiDAR and images captured by high-speed cameras, a 3D model of the dust field caused by vehicle wheels is reconstructed using computer vision algorithms and 3D reconstruction technology, intuitively displaying the distribution pattern, diffusion range, and trajectory of the dust. First, multiple cameras of the high-speed camera module are precisely calibrated to determine their positions and orientations in the world coordinate system. Next, feature points are extracted from the image sequences captured by multiple cameras. Feature extraction algorithms such as Scale Invariant Feature Transform (SIFT), Speed-Up Robust Feature Transform (SURF), or Oriented Fast and Rotated BRIEF (ORB) are used to detect unique and stable feature points in each frame. These feature points maintain good recognizability under different viewing angles and lighting conditions. After feature point extraction, feature matching is performed, and the RANSAC (Random Sample Consensus) algorithm is used to remove mismatched points, resulting in a set of feature points corresponding to the same spatial point in different images. Based on the matched feature point pairs, the 3D coordinates of the spatial points are calculated using triangulation principles. The large number of calculated 3D spatial points are then combined into point cloud data. Point cloud data initially depicted the spatial distribution of lunar dust particles in the vehicle dust field at different times. The point cloud data was then filtered to remove noise points. A statistical filtering method was used to calculate the statistical distance between each point and its neighbors; points deviating significantly from the statistical distribution were considered noise and removed. Finally, a surface reconstruction algorithm was used to transform the filtered point cloud data into a continuous surface model.

[0042] This method's 3D reconstruction is tailored to the characteristics of the dust field generated by the lunar rover's movement on the lunar surface. It considers the influence of the low gravity and vacuum environment on the lunar dust's trajectory, as well as the dynamic changes in the dust field caused by the interaction between the rover's wheels and the lunar regolith. This allows for a more accurate reflection of the 3D distribution and movement of lunar dust in this specific scenario. During 3D reconstruction, in addition to image data provided by a high-speed camera module to acquire information on the position and trajectory of lunar dust particles, this method also incorporates information on the particle size distribution and concentration of lunar dust obtained from a laser measurement module. This multi-source data complements each other, enabling a more comprehensive description of the dust field's characteristics during 3D reconstruction.

[0043] Dust Characteristics Analysis: Combining particle counter and lidar data, this study comprehensively analyzes the particle concentration, particle size distribution, and relationship with environmental factors of dust. Particle counter data reflects the concentration changes and particle size distribution patterns of dust particles. Based on lidar module measurement data, spatial interpolation is performed on the concentration and particle size data at discrete measurement points to generate continuous concentration and particle size field distribution maps. By calculating the gradient distribution of the concentration field, the region with the highest dust concentration can be identified. Analyzing the spatial differences in the particle size field and statistically analyzing the proportion of lunar dust particle sizes in different regions (such as near-surface and high-altitude regions), combined with lunar dust trajectory and velocity data, the external forces acting on the lunar dust (such as lunar gravity, wheel thrust, and inter-particle collision forces) can be inverted, and the force differences for particles of different sizes can be calculated, explaining why finer particles are more easily suspended. This method, through multi-module data fusion, achieves multi-dimensional analysis of spatial distribution, temporal dynamics, and physical characteristic correlations, systematically revealing the overall characteristics of the dust field.

[0044] Data Fusion and Modeling: The various data sets mentioned above are fused to establish a dynamic model of the lunar rover's wheel dust field. This model can reflect the spatiotemporal changes of the dust field in real time, predict the diffusion trend and impact range of dust, and provide a basis for decision-making regarding the lunar rover's driving planning and mission scheduling. Key features are extracted from the data of each module: the high-speed camera module extracts the motion trajectory features (such as position, velocity vector, and acceleration) and morphological features (such as particle projected area) of lunar dust particles; the laser module extracts particle size distribution features (such as average particle size and particle size distribution histogram) and concentration features (such as the number of particles per unit volume). By calculating the similarity between features (such as associating the motion features of particles on a certain trajectory with the particle size and concentration features of that point when the particles are near the laser measurement point), a multi-factor feature mapping relationship is established. A filtering algorithm is used to optimize the associated features, and the particle size estimation error in the high-speed camera is corrected by combining the particle size data from laser scattering. At the same time, duplicate or conflicting features are removed (when the concentration measured by the laser at the same spatial point deviates from the concentration converted by the high-speed camera counting beyond a threshold, the laser measurement value is used). This paper utilizes data to construct a multi-input neural network model, taking features such as trajectory, particle size, concentration, and charge after feature layer fusion as inputs, and outputting comprehensive parameters of the dust field. Existing data fusion technologies are mostly limited to a single dimension, such as fusing only spatial data from optical images and lidar, or only fusing concentration measurement data from different sensors. This method, however, achieves multi-dimensional data fusion across both spatiotemporal and spatial scales.

[0045] Results visualization and presentation: The observation results are visualized in an intuitive way. The dynamic changes of the vehicle dust field can be clearly understood through three-dimensional graphics, which facilitates subsequent research and analysis.

[0046] The observation device for in-situ observation of dust fields from lunar rover wheels includes multiple observation components and their supporting structures mounted on the lunar rover, including: The high-speed camera 1 is installed in dedicated holes on both sides of the front dust shield of the lunar rover, arranged symmetrically, to capture the dynamic movement of lunar dust particles within the wheel area at a high frame rate. The high-speed camera is mounted with a mechanical mounting bracket to ensure that its field of view is stably pointed at the dust-generating area, and it has a dustproof and shockproof encapsulation structure to adapt to the low pressure and vibration environment of the lunar surface.

[0047] The particle counting sensor 2 is located on the rear side of the lunar rover, near both sides of the wheels, and is fixed to the rover structure via an extension bracket. This sensor is used to collect real-time data on the concentration and size distribution of suspended lunar dust particles in the air near the wheels. Its air inlet faces the dust-generating area and is equipped with a primary filter and protective cover to prevent large particles from damaging the internal sensitive components.

[0048] The lidar 3 is positioned slightly rear-centrally on the vehicle body and mounted on a platform via a rotating bracket. It can perform laser scanning of the terrain and suspended dust particles around the wheels to acquire three-dimensional point cloud data. The lidar has a certain pitch angle adjustment function, which can adjust the observation area in real time according to the driving direction.

[0049] The lunar rover's dust shield 4 is positioned above the outer sides of the tires around the rover's perimeter, forming a physical barrier against dust splash. Its structure includes pre-drilled holes for mounting high-speed cameras. Made of lightweight, high-strength materials, the dust shield combines structural protection with equipment support, effectively reducing interference from splashed lunar dust on sensitive equipment.

[0050] The global camera 5 is mounted on the upper front of the vehicle body at a relatively high position and is fixed by a universal adjustment device. The camera's field of view covers the entire wheel and dust field area, and is used to acquire wide-area image data to assist in recording the overall dust field morphology and background modeling, and to provide contextual information support for local images.

[0051] In this embodiment, the calculation method is as follows: Perspective projection formula in camera calibration: The camera projects a 3D spatial point (X,Y,Z) onto a 2D image point (u,v), using the following formula:

[0052] Where K is the camera intrinsic parameter matrix, R is the rotation matrix, and t is the translation vector, used to describe the camera's attitude and position in the coordinate system.

[0053] Formula for calculating the coordinates of a point in space using triangulation: Given the projection matrices of two cameras P 1 and P2, and the corresponding image points x 1 and x 2. Spatial point X satisfies and Through calculation, we can obtain:

[0054] in P + The pseudo-inverse of the projection matrix Statistical filtering formula for removing noise points Calculate each point in the point cloud and its relation to k The average distance between the nearest neighbors μ and standard deviation σ When the average distance between a point and its neighboring points d satisfy , ( g When the threshold coefficient (usually 2 or 3) is used, it is determined to be a noise point, and the formula is:

[0055] in d i For this point and the first i The distance between the nearest neighbors.

[0056] Concentration field gradient:

[0057] The magnitude of the gradient reflects the rate of concentration change, and its direction is the direction of the fastest increase in concentration.

[0058] Spatial interpolation formula: Used to generate continuous concentration field and particle size field distribution maps, at a certain point x The estimated value of 0

[0059] in, Z(x i ) For known measurement points x i Concentration or particle size value, λi Let be the weighting coefficient, satisfying

[0060] The weighting coefficients are calculated using a semivariogram. The formula is:

[0061] h The distance between sample points N (h) The distance ish The number of sample point pairs.

[0062] Feature similarity formula: Calculate the similarity between the motion features extracted by the high-speed camera module and the features extracted by the laser module, for two feature vectors. and cosine similarity sim for:

[0063] The closer the value is to 1, the more similar the features are. Multi-sensor data fusion formula: Let the lidar measurement value C 1. High-speed camera concentration conversion C 2, with weights respectively ω 1 and ω 2( ω 1+ ω 2=1), the concentration C after fusion is:

[0064] when ( (for deviation thresholds), take ω 1=1、 ω 2=0.

[0065] The above description of several specific embodiments further details the technical solution provided by the present invention in order to highlight the advantages and benefits of the technical solution provided by the present invention. However, the above-described specific embodiments are not intended to limit the present invention. Any reasonable modifications and improvements to the present invention, combinations of embodiments, and equivalent substitutions based on the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for in-situ dynamic observation of dust fields from lunar rover wheels, characterized in that, include: The steps involve synchronously processing, 3D reconstructing, fusing analysis and modeling image data, laser point cloud data and particle concentration and size data, and outputting dynamic parameters of the dust field. in, Fusion analysis includes aligning image point clouds and laser point clouds in space and unifying particle size and concentration feature estimation; Specifically, the high-speed camera module extracts the motion trajectory and morphological features of lunar dust particles, while the laser module extracts the particle size distribution and concentration features. Multi-factor feature mapping relationships are established by calculating the similarity between features. The processing of particle size and concentration characteristics includes generating concentration and particle size fields using spatial interpolation methods, and extracting concentration gradients and regional features; Specifically, based on the particle size and concentration data output by the particle counter, and combined with the corresponding measurement points in the laser point cloud, spatial interpolation is performed to generate a continuous concentration field and particle size field.

2. The method for in-situ dynamic observation of lunar rover wheel dust field according to claim 1, characterized in that, Image data processing includes image enhancement, feature extraction, and feature matching of image sequences, and the generation of three-dimensional point clouds using multi-view triangulation.

3. The method for in-situ dynamic observation of lunar rover wheel dust field according to claim 1, characterized in that, Statistical filtering is applied to the generated point cloud to remove noise points and construct a continuous surface model.

4. A dynamic in-situ observation device for dust field from lunar rover wheels, characterized in that, include: A module that simultaneously processes, reconstructs, fuses, analyzes, and models image data, laser point cloud data, and particle concentration and size data, and outputs dynamic parameters of the dust field. in, Fusion analysis includes aligning image point clouds and laser point clouds in space and unifying particle size and concentration feature estimation; Specifically, the high-speed camera module extracts the motion trajectory and morphological features of lunar dust particles, while the laser module extracts the particle size distribution and concentration features. Multi-factor feature mapping relationships are established by calculating the similarity between features. The processing of particle size and concentration characteristics includes generating concentration and particle size fields using spatial interpolation methods, and extracting concentration gradients and regional features; Specifically, based on the particle size and concentration data output by the particle counter, and combined with the corresponding measurement points in the laser point cloud, spatial interpolation is performed to generate a continuous concentration field and particle size field.

5. A dynamic in-situ observation system for dust fields from lunar rover wheels, characterized in that, To implement the method of claim 1, the method comprises: A high-speed camera mounted on the dust baffle is used to capture dynamic images of dust. A particle counting sensor installed on the side of the vehicle body is used to detect the concentration and size of dust particles; A lidar unit mounted on the vehicle body is used to acquire spatial distribution information of dust. A global camera mounted high on the vehicle body is used to capture overall images of the dust field.

6. A computer storage medium for storing computer programs, characterized in that, When the computer program is read by the computer, the computer executes the method of claim 1.

7. A computer, comprising a processor and a storage medium, characterized in that, When the processor reads the computer program stored in the storage medium, the computer executes the method of claim 1.

8. A computer program product, as a computer program, is characterized by: When the computer program is executed, it implements the method of claim 1.

Images