1066 results about "Lidar point cloud" patented technology

Disclosed are a high-performance loosely-coupled multi-modal data fusion system for a smart driving environmental perception system and a vehicle-mounted device, comprising: a fusion detection model based on a modality-independent feature interaction strategy, which is configured for converting a LiDAR point cloud, a camera image, and a millimeter-wave radar point cloud into a unified bird's-eye view representation, and performing multi-modal fusion; and a fusion tracking model based on a motion-appearance feature cascaded coupling data association strategy, which is configured for performing subsequent trajectory tracking and matching according to multi-modal fusion feature information. A VoD data set and a K-Radar data set are selected for training, verifying, and testing the comprehensive performance of the models, and a TensorRT accelerated inference model is applied, then quantized, and deployed to a vehicle-mounted computational testing platform. The present invention is compatible with mainstream sensor deployment solutions, and achieves the efficient complementary fusion of multi-source heterogeneous sensor information, significantly improving the reliability, accuracy, and adaptability of vehicle-mounted perception systems, thereby effectively responding to extreme operating conditions such as complex traffic scenarios and inclement weather.

The invention relates to a tower crane operation control system based on complex scene three-dimensional real-time modeling. According to the system, a lifting hook is coarsely positioned through a lifting hook positioning and state sensing unit, a real-time position is positioned by combining a laser radar point cloud clustering algorithm with historical pose data, and visual tracking is synchronously performed by means of a tower top camera AI; converting the real-time point cloud data into a 3D voxel grid map, generating a global path by using a 3DA algorithm, and outputting a hoisting track after smooth processing and track optimization; establishing a sling-lifting hook double-pendulum dynamic model, predicting a state sequence based on a model prediction control algorithm, and adjusting a control signal through a feedforward compensation item and a feedback correction item; and the man-machine interaction and monitoring unit is used for displaying the cantilever angle, the lifting hook height and the three-dimensional map of the tower crane in real time and remotely intervening the operation state of the tower crane. According to the system, multi-source data are fused to construct a high-precision three-dimensional map, lifting hook positioning and full-view tracking are achieved, and lifting safety and trajectory tracking precision are improved through path planning and dynamics control.

The invention relates to the technical field of ground mobile unmanned equipment control, and discloses a ground mobile unmanned equipment autonomous obstacle avoidance control system optimized by artificial intelligence. The system comprises an environment perception layer, a bimodal risk assessment layer, a dynamic decision-making layer, a trajectory optimization layer and a feedback optimization layer. The environment sensing layer adopts a retina fovea centralis imitating mechanism to perform non-uniform sampling on laser radar point cloud data to generate dynamic point cloud partitions; the bimodal risk assessment layer fuses two types of radar data to generate static and dynamic obstacle risk assessment diagrams; the dynamic decision-making layer establishes space-time mapping and generates an obstacle confidence coefficient matrix through a graph neural network; the trajectory optimization layer converts the matrix into a control parameter based on a multi-objective evolutionary algorithm, and issues the control parameter through a time-sensitive network protocol; and the feedback optimization layer monitors environment change, calculates deviation, generates an effectiveness index, and dynamically adjusts a point cloud acquisition strategy until the index is optimal. According to the system, the autonomous obstacle avoidance capability and adaptability of the ground mobile unmanned equipment in a complex environment are enhanced.

The invention relates to the technical field of risk prediction, in particular to a collision risk prediction method of a low-altitude aircraft in high-density flight in a complex environment, which comprises the following steps: acquiring a point cloud through a three-dimensional laser radar, clustering to generate an obstacle trajectory, matching the trajectory to identify a disturbance characteristic segment, calculating a deviation angle by combining a path vector to generate a disturbance frequency graph, and calculating the collision risk of the low-altitude aircraft. And extracting a parameter modeling dynamic safety interval, and fusing multiple factors to evaluate a collision risk level. According to the method, obstacle trajectory topology is constructed through combination of three-dimensional laser radar point cloud time window division and density clustering, sudden change features are identified through trajectory similarity matching, a Gaussian kernel dynamic safety envelope is generated through combination of course offset statistics and included angle operation, and a self-matching threshold value is established through normalization parameters and radial basis weighting. The method improves the high-density flight collision prediction precision, enhances the weather and obstacle heterogeneity matching capability, reduces the misjudgment early warning delay, and achieves the multi-variable flight situation collaborative analysis.

The invention relates to the technical field of land resource monitoring, in particular to a land resource dynamic monitoring and early warning method and system based on multi-source remote sensing data fusion, and the method comprises the steps: employing an unmanned plane to periodically collect optical images, SAR echoes and LiDAR point clouds, constructing a ground three-dimensional digital model, and carrying out the land parcel division; performing fusion to form a multi-dimensional feature vector, establishing an LSTM land parcel feature evolution model, and predicting a change rate interval of each feature in a current period based on a historical sequence; constructing a time sequence difference change detection algorithm, calculating a land parcel change rate, and screening potential abnormal land parcels by taking a prediction interval as an anomaly judgment threshold value; a double-branch convolutional neural network is adopted to identify crop states, growth stages and construction violation behaviors, abnormity is judged and determined, and confidence is given; spatial clustering is carried out on determined abnormal land parcels, accurate boundaries are obtained in combination with a three-dimensional model, multi-level early warning information is generated, and the decision-making efficiency and response speed of land resource monitoring are improved.

The invention discloses a landslide mass monitoring method based on an unmanned aerial vehicle laser radar. The landslide mass monitoring method comprises the following steps: S1, constructing an exposed earth surface point cloud model of a first time phase; s2, constructing an exposed earth surface point cloud model of a second time phase; s3, establishing a standard point cloud data set; s4, constructing a multi-dimensional time sequence feature vector sequence; s5, introducing a fusion prediction model of the ARIMA and the neural network, and outputting a final prediction sequence of the deformation trend of the landslide mass; and S6, constructing a disaster chain propagation model, outputting a landslide early warning grade according to a preset grading threshold value, and performing real-time early warning information release and response linkage. According to the invention, through fusion of laser radar point cloud registration analysis and ARIMA-neural network prediction modeling, high-precision dynamic monitoring and early warning of tiny deformation of the landslide mass are realized, and the method is suitable for landslide early identification and disaster emergency response scenes in a complex terrain environment.

The invention relates to the technical field of surveying and mapping, and discloses a laser radar point cloud data processing surveying and mapping method and system based on deep learning, and the method comprises the steps: obtaining original point cloud data collected by a laser radar, carrying out the preprocessing of the original point cloud data, and obtaining the preprocessed point cloud data; taking a U-Net network as a basic framework, and introducing residual connection and a multi-scale feature fusion mechanism to carry out de-noising processing on the preprocessed point cloud data to obtain de-noised point cloud data; inputting the denoised point cloud data into a Point CNN-GAT model to extract local geometric features and a global topological relation, introducing a dynamic cavity convolution module, and adaptively adjusting the sampling range of a convolution kernel; registering the comprehensive feature vector output by the Point CNN-GAT model, and generating a corresponding surveying and mapping result according to the registered point cloud data; according to the invention, the time cost of data circulation and manual intervention is reduced, and the overall processing efficiency is improved.

The invention provides a multi-modal fusion and semantic enhancement train positioning method and system, and belongs to the technical field of rail transit, and the method comprises the steps: carrying out the time-space alignment of data, obtaining a dense point cloud, constructing a dense semantic point cloud, and dynamically estimating the confidence coefficient weight of each type of sensors; a set residual error and a Manhattan structure constraint residual error of a plane are constructed, laser radar point cloud parameters are obtained, and visual projection constraints are constructed at the same time; constructing a comprehensive degradation scoring function to carry out degradation judgment on the current environment; when the degradation result is yes, introducing a structure and motion information independent of an external environment, maintaining trajectory estimation, and constructing a compensation constraint; introducing a prior semantic constraint and a large model semantic factor constraint; and constructing a global optimization objective function, dynamically adjusting the weight of each modal factor, obtaining an optimal estimation state, and outputting a high-precision train positioning result. According to the method, high-precision and robust track estimation in an extreme scene is realized, so that the continuity, safety and intelligence of train positioning are guaranteed.

The embodiment of the invention discloses a multi-sensor fusion anti-degradation SLAM mapping method and system. The method can effectively solve the problem of pose drift of a robot in a mapping process in structure degradation environments such as an indoor long corridor, constructs a globally consistent three-dimensional point cloud map and a robot trajectory, and comprises the following steps: realizing depth coupling of an IMU and a wheel speedometer based on extended Kalman filtering, and generating high-frequency pose prediction; denoising, down-sampling and motion distortion correction are carried out on the 4D laser radar point cloud, and the normal vector and intensity characteristics of the point cloud are extracted; a normal vector and intensity feature enhanced scanning matching algorithm is adopted, and a target function is optimized through a multi-feature weight, so that the matching precision in a degradation scene is improved; loopback detection is realized through candidate key frame screening and geometric registration verification, and a closed-loop constraint is incorporated into a factor graph for global correction; and finally, incrementally updating the global point cloud map and carrying out consistency optimization, and outputting a robust three-dimensional point cloud map and a high-precision robot track.

The invention discloses a method for judging an RTK abnormal value in an automatic driving integrated navigation system, and relates to the technical field of automatic driving high-precision integrated navigation, and the method comprises the steps: collecting RTK observation data, inertial measurement unit data, a wheel speed pulse signal and LiDAR point cloud data, carrying out timestamp alignment and coordinate system unification, and generating a fusion data value; calculating a carrier-to-noise ratio weight signal quality index of the satellite through the fused data value, and generating a signal quality report; and combining the signal quality report, the current satellite geometric accuracy factor and the vehicle motion acceleration, calculating a residual threshold, constructing a coriolis force compensated double-integral prediction model by using inertial measurement unit data, and predicting the position of the vehicle at the current moment. According to the method, multi-dimensional features such as satellite signal quality, vehicle motion state and residual analysis are fused through multi-source decision, and the anomaly detection capability of complex scenes such as urban canyons is effectively improved.

The invention discloses a self-adaptive 4D Gaussian splashing high-precision three-dimensional reconstruction system and method, and belongs to the technical field of computer graphic processing. The invention aims to realize high-precision automatic registration of multi-source heterogeneous data and improve the calculation efficiency. The method comprises the following steps: collecting multi-source heterogeneous data; constructing a multi-modal fusion registration method, which comprises the following steps: combining satellite image data and low-altitude oblique photography data to realize spatial distribution geometric coarse registration, fusing low-altitude laser radar point cloud data and ground acquisition vehicle laser radar point cloud data to realize luminosity fine registration, establishing semantic features to assist registration, and obtaining registered multi-source data; initializing a 4D Gaussian primitive and executing adaptive splashing reconstruction to obtain an optimized 4D Gaussian splashing model; designing a cloud edge cooperative computing architecture oriented to 4D Gaussian splash reconstruction, and performing distributed parallel processing on the obtained optimized 4D Gaussian splash model; and executing quality evaluation and adaptive optimization, and outputting a final adaptive 4D Gaussian splash model.

The invention discloses a laser radar and multispectral polarization imaging fused target identification system and method. The system comprises a sensor configuration and data preprocessing module, a cross-modal feature extraction and fusion module and a multi-task output and optimization module. The sensor configuration and data preprocessing module performs time-space synchronization processing on the collected original optical signals and laser signals in the environment to obtain multispectral image data and laser radar point cloud data; the cross-modal feature extraction and fusion module performs feature extraction and fusion enhancement processing on the multispectral image data and the laser radar point cloud data, and outputs high-dimensional semantic enhancement point cloud representation containing image semantics and point cloud geometry; and the multi-task output and optimization module processes the high-dimensional semantic enhanced point cloud representation and outputs a three-dimensional target recognition result, target speed information and a pixel-level depth map. Through module design and data processing, the defects in the prior art are overcome, and the accuracy of target recognition is improved.

The invention discloses a geotechnical engineering slope deformation monitoring method and system. The method comprises the following steps: acquiring multi-modal image data, enhancing the weak deformation area contrast of the multi-modal image data by using a CLAHE algorithm, extracting gradient edge information through a Sobel operator, and filtering laser radar point cloud noise based on VMD; capturing a bidirectional dependency relationship of a time sequence based on a BiLSTM bidirectional long short-term memory network, and dynamically allocating weights of different time steps and spatial positions through an attention mechanism; optimizing hyper-parameters of the hybrid prediction model by using a GJO Golden Litsea Optimization algorithm to obtain a target hybrid prediction model; and inputting the feature image data into the target hybrid prediction model for prediction, generating a risk probability thermodynamic diagram, and positioning the position of the potential slip crack surface of the slope according to the risk probability thermodynamic diagram. And reliability and accuracy of geotechnical engineering slope image data analysis are improved.

The invention discloses a fine individual tree segmentation method fusing air-based laser radar point clouds and ground-based laser radar point clouds, and aims to solve the challenges that the data volume of combined point clouds is large, crown staggering is complex in a dense forest environment, and part of tree trunk point clouds are lost due to shielding, and the method combines an individual tree segmentation strategy from top to bottom and an individual tree segmentation strategy from bottom to top. A weighted height-density map is adopted to carry out individual tree positioning, a mark control watershed algorithm is combined to realize preliminary tree crown region extraction, then multi-label local graph segmentation optimization is introduced, segmentation boundaries are corrected, an under-segmentation unit is optimized by adopting a shortest path, and a complete segmentation framework is formed. On the basis of airborne and foundation point clouds of six actually measured sample plots, the method is verified after registration and combination. The advantages of airborne point cloud and foundation point cloud are combined, and the proposed individual tree segmentation method provides technical support for refined forest modeling and ecological monitoring.

The invention discloses a railway tunnel foreign matter detection method based on radar point cloud and RGB image fusion. The method comprises the following steps: acquiring a laser radar point cloud and an RGB image in a railway tunnel; obtaining an RGB image by using an adversarial network model GAN to judge the abnormal occurrence state of the foreign matter in the railway tunnel, and if the abnormal state of the foreign matter occurs, projecting the obtained laser radar point cloud to an image plane to obtain a dense depth map; processing through a sliding window to obtain data slices which are approximately square; and carrying out data feature extraction, fusion and foreign object target detection on the data slices of the RGB image and the dense depth map by using a fusion target detection network based on the OfficientNet-FPN, and outputting a foreign object target information detection result containing a target category, a position, a confidence coefficient and a distance. According to the railway tunnel foreign matter detection method based on radar point cloud and RGB image fusion, high precision, high efficiency and high reliability of railway tunnel foreign matter detection are realized.

The invention relates to the technical field of autonomous navigation of inspection robots, and provides a field inspection robot autonomous navigation and path planning method based on multi-modal sensing. The method comprises the following steps: collecting a laser radar point cloud, a camera image, inertial measurement and positioning data, obtaining an environment semantic feature set through spatio-temporal feature fusion, and generating a semantic occupation map; inputting the current position of the robot, the target point and the semantic occupation map into a trajectory generation network to obtain candidate trajectories meeting obstacle avoidance and path smoothness constraints, and completing task sorting and trajectory splicing in combination with inspection task points to form a global path; in the operation process, the reinforcement learning control model adjusts the linear speed and the angular speed in real time, and path tracking and dynamic obstacle avoidance are achieved. The navigation precision and the operation safety of the inspection robot in the complex field area are improved.

According to the permanent basic farmland quality intelligent monitoring system and method provided by the invention, accurate monitoring and dynamic management are realized through multi-module cooperation, the terrain gradient and canopy height are calculated based on LiDAR point cloud, the complexity index is generated in combination with meteorological data, and the A * algorithm is improved to optimize the unmanned aerial vehicle inspection path. The method comprises the following steps: extracting hyperspectral data to analyze soil and crop states, calibrating an electrode response curve, calculating a bio-availability index in combination with temperature and humidity, dynamically adjusting an early warning threshold value, identifying abnormal data by adopting an isolated forest algorithm and carrying out grading processing, and constructing a space-time cube by adopting space-time Kriging interpolation and CNMF algorithm and fusing multi-source data. According to the method, indexes such as ecological protection grades and economic values are integrated to generate a graded soil quality index model, a disease thermodynamic diagram is generated, key data are stored in a chaining mode, and farmland quality real-time monitoring, risk early warning and ecological economic balance decision making are achieved through multi-source heterogeneous data fusion, dynamic weight optimization and block chain technologies.

The invention provides a multi-source data fusion self-positioning method and system. The method comprises the following steps: acquiring GPS positioning data, visual image data and laser radar point cloud data of an unmanned aerial vehicle; carrying out noise reduction preprocessing on the GPS positioning data by adopting an unscented Kalman filtering algorithm; performing sensor joint calibration based on a visual image and a laser radar point cloud, and establishing a geometric mapping relation between a camera coordinate system and a world coordinate system by retrieving a preset high-precision tower ledger library and solving a PnP problem; inputting the image target detection data, the GPS state estimation value and the geometric mapping data into a pre-trained auto-encoder regression network; anti-interference potential features are extracted through an encoder of the network, and a target position estimation value of the target space position of the unmanned aerial vehicle is output through a regression head. According to the invention, the problem of positioning drift caused by strong electromagnetic interference and complex landform in electric power inspection is effectively solved.

The invention provides a traffic abnormity and congestion cause analysis method and system based on a knowledge graph, and relates to the technical field of intelligent traffic perception, and the method comprises the steps: carrying out the target detection through employing preprocessed laser radar point cloud data, recognizing traffic participants, carrying out the continuous frame tracking of the traffic participants, and extracting the motion track and behavior characteristics in an event dimension; identifying a behavior event of the traffic target based on the motion trail and the behavior characteristics, binding the identified behavior event of the traffic target to a corresponding target entity, and performing target behavior modeling to form standardized structured information; based on structured information of a traffic target, an intersection-oriented traffic state knowledge graph is constructed, a rule-based abnormal event judgment module is utilized to perform semantic analysis on behavior and event nodes in the traffic state knowledge graph, abnormal traffic events are identified, and potential causes causing traffic congestion are traced. According to the invention, refined understanding and active perception of the intersection traffic state can be realized.

The invention belongs to the technical field of robot navigation, and particularly relates to a humanoid robot navigation method based on visual semantic segmentation and radar obstacle detection, which comprises the following steps: synchronously acquiring an RGB image and a depth image of a current environment of a humanoid robot by using a visual sensor, and acquiring point cloud data by using a laser radar; performing semantic segmentation on the preprocessed RGB image; radar point cloud obstacle detection; fusing the semantic segmentation map and the laser point cloud map, introducing a Bayesian decision to judge whether the laser point cloud map is passable or not, and then calculating a fusion cost value to obtain a fusion cost map; adopting an RRT * / TEB algorithm to output an optimal path; using a nonlinear optimization solver to generate a foothold sequence accurate to each step; according to the method, a semantic-geometric two-dimensional navigation cost model is constructed; bayesian reasoning is deeply bound with a navigation scene, so that the navigation adaptability of an unstructured environment is improved; the navigation method is suitable for the humanoid robot, and is low in energy consumption, low in navigation deviation and high in safety.

The invention discloses a low-slow small target detection and trajectory prediction tracking method based on a laser radar, and the method comprises the steps: firstly collecting the point cloud data of the laser radar, and carrying out the preprocessing of spatial modeling and coordinate transformation of the point cloud data of the laser radar; performing significance screening; based on distance partition driving, Pilllar construction and coding are carried out; constructing a deep learning detection network; based on Anchor design and a matching strategy, carrying out size adaptation on a weak target in the air in the fused features; and performing time sequence prediction and observation updating on the target state based on an extended Kalman filter (EKF), and completing low-slow small target detection and trajectory prediction tracking. The method can maintain the high precision advantage of the laser radar, improves the recognition capability of the laser radar on weak-reflection, small-size and irregular-motion targets, has high robustness and environment adaptability, and achieves the stable and precise sensing and continuous tracking of low, slow and small flight targets.

The invention relates to an earthwork balance calculation method, system and device and a storage medium, and relates to the technical field of earthwork balance design. The method comprises the steps of performing fusion processing on oblique photography data and laser radar point cloud data to generate a live-action three-dimensional model; performing grid division on the live-action three-dimensional model to obtain a grid model; based on the gridding model, defining a decision variable set including the design elevation of each grid point, the overall slope of the site, the slope height of each slope unit and the earthwork allocation amount, and constructing a comprehensive objective function including an earthwork balance objective, an economical objective and a safety objective; and under the engineering constraint condition, performing multi-objective optimization solution on the comprehensive objective function to obtain a Pareto optimal solution set, and selecting a final implementation scheme from the Pareto optimal solution set. According to the method, the problem of topographic data missing of the dense vegetation region is effectively solved, the limitation of traditional single-target optimization is broken through, and the maximization of comprehensive benefits is realized.

The invention relates to the technical field of multi-unmanned-aerial-vehicle cooperative operation and three-dimensional modeling, in particular to a multi-unmanned-aerial-vehicle cooperative three-dimensional rapid modeling method for a highway accident scene, and the method comprises the steps: generating a three-dimensional grid map of an accident area through the scanning of a millimeter-wave radar by a main control unmanned aerial vehicle; subareas are divided according to a load balancing strategy and are distributed to slave unmanned aerial vehicles, the slave unmanned aerial vehicles traverse grids along a snake-shaped track, laser radar point clouds and five-view-angle images are synchronously collected, data are bound through double time stamps and space coordinates, the point clouds are preprocessed through edge computing nodes, and the point clouds are stored in a database; the master control unmanned aerial vehicle evaluates quality based on density standard deviation and overlapping matching degree and instructs to reacquire, performs high-precision Poisson reconstruction on an accident core area, performs voxelization processing on a peripheral area, maps image textures, optimizes vehicle deformation details, simplifies a model and retains key element precision, and finally performs data processing. License plate coordinates, a scattered object thermodynamic diagram and an emergency lane occupation state are automatically marked, a visual model is generated, and rapid and accurate restoration of an accident scene is realized.

In various examples, self-supervised learning may be used to pre-train an encoder network of a masked prediction model to reconstruct masked regions of an input representation of 3D detections such as LiDAR point cloud(s). Spatial and / or temporal masking may be applied to a projected representation of 3D detections (e.g., a two-dimensional (2D) projection image), and the masked prediction model (e.g., a masked auto-encoder or joint-embedding predictive architecture) may be used to reconstruct a representation of the masked regions (e.g., reflection characteristic(s) stored in corresponding pixels or cells of the projected representation, a latent representation of the reflection characteristic(s)) during iterations of self-supervised learning. As such, the pre-trained encoder network of the masked prediction model may be used as a foundation model and fine-tuned with a task-specific output head or its pre-trained weights may be used to initialize a task-specific model.

The invention belongs to the technical field of laser radar-camera panoramic segmentation, and particularly relates to a multi-mode panoramic segmentation method for multi-view space-time alignment and implicit feature interaction, and the method is executed by a multi-mode panoramic segmentation network, and comprises the steps: S1, obtaining laser radar point cloud data and multi-view camera image data in the same scene; s2, performing double-branch feature coding on the laser radar point cloud data; performing multi-scale image feature extraction on the camera image data; s3, generating false point cloud features with geometric perception capability; s4, generating semantic pixel features; s5, performing implicit fusion on the pseudo point cloud features generated in the S3 and the semantic pixel features generated in the S4 to obtain cross-modal fusion features; and S6, based on the cross-modal fusion features obtained in the S5, generating a unified panoramic segmentation result containing semantic tags and instance IDs. The method can effectively improve the robustness and precision of multi-mode panoramic segmentation in a complex urban environment.

The invention belongs to the field of target identification, and relates to a multi-modal target identification method and device for an unmanned aerial vehicle, electronic equipment and a storage medium, and the method comprises the steps: collecting multi-modal data, and carrying out the preprocessing of the multi-modal data, the multi-modal data comprises visible light image data, infrared thermal imaging data, laser radar point cloud data and synthetic aperture radar data; carrying out hierarchical cross-modal feature extraction on the basis of the preprocessed multi-modal data; multi-scale cross-modal feature fusion based on attention guidance is carried out; predicting the position and category of each target in the image based on the fused multi-scale features; assigning a unique ID to each target by associating detection results in continuous frames to form a motion track, and correcting an identification result of the current frame by using spatio-temporal context information; and lightweight model optimization and embedded real-time deployment are carried out. The problems of target shielding and losing can be solved, current frame identification can be further optimized, and tracking continuity and accuracy are ensured.

The invention provides a multi-target tracking information enhanced laser radar visual inertia high-precision positioning method, and belongs to the technical field of environmental perception and navigation positioning. The positioning method comprises the following steps: acquiring measurement data of a laser radar, a camera and an IMU (Inertial Measurement Unit), and performing data preprocessing, including laser radar point cloud data preprocessing, image data preprocessing and IMU pre-integration; based on the preprocessed data, establishing a target comprehensive association scale fusing a laser radar and a visual multi-modal detection bounding box and geometric features, and associating multi-modal target data; the continuous and reliable multi-target tracking and self-localization performance is realized by combining the multi-modal detection bounding box and the geometrical characteristics of the target and the IMU pre-integration measurement optimization carrier pose and dynamic target trajectory.

The invention relates to the technical field of geographic information engineering, in particular to a surveying and mapping geographic system based on a remote sensing technology, which is characterized in that a multi-modal remote sensing data acquisition module synchronously acquires satellite remote sensing images, aviation LiDAR point cloud, unmanned aerial vehicle oblique photography and ground Internet of Things sensing data; the data fusion collaboration module carries out geographic coordinate calibration and position coding, and generates a surveying and mapping data set with unified precision in combination with an improved Mamba fusion model; the dynamic terrain high-precision inversion module performs three-dimensional terrain reconstruction, and completes terrain deformation monitoring in combination with an InSAR interference measurement technology and a time sequence analysis algorithm; the real-time error correction module constructs a BP neural network optimized by an ISSA algorithm, and dynamically corrects a terrain three-dimensional model and a terrain deformation result; the intelligent visual application module constructs a digital twinborn simulation model and automatically generates a standardized surveying and mapping result report. Therefore, the problems of weak multi-source data fusion, low terrain inversion precision and the like in the prior art are solved.

The invention relates to a laser radar camera calibration method and device, and a medium. The method comprises the steps: collecting multi-frame laser radar point cloud data and synchronous corresponding image data in a construction scene; carrying out multi-frame point cloud fusion and dense reconstruction to generate a laser dense point cloud; visual sparse point cloud reconstruction is carried out, and cross-modal scale unification and space alignment are carried out on an initial visual point cloud obtained through reconstruction and the laser dense point cloud; generating a visual dense point cloud through three-dimensional Gaussian splashing; and performing registration on the laser dense point cloud and the visual dense point cloud by adopting a point-to-line iterative nearest point algorithm, and performing calculation to obtain an external parameter calibration matrix of the camera and the laser radar. Compared with the prior art, the method has the advantages of high precision, low cost, high stability and the like.

The invention discloses a multi-sensor fusion processing method, perception method and equipment based on LiDAR point cloud pseudo image conversion, a multi-view projection strategy is adopted to convert a preprocessed LiDAR point cloud into an aerial view, a front view and a side view, the aerial view, the front view and the side view are subjected to feature coding and then fused through a SENet attention mechanism to generate a multi-view fusion pseudo image, and a large amount of space information is reserved. Meanwhile, in order to effectively improve the multi-sensor feature fusion efficiency, after a visual image is preprocessed, an improved CNN network is adopted to extract the features of a LiDAR pseudo image and the visual image, feature fusion is achieved through a cross-modal attention mechanism, attention weights are generated by calculating a similarity matrix, and the feature fusion efficiency is improved. And after the features are enhanced, the features are fused in a channel splicing and element-level adding mode. The method can effectively improve the precision and robustness of automatic driving environment perception, enhances the performance in a complex scene, and is suitable for tasks such as target detection and semantic segmentation in automatic driving.