34 results about "Vector map" patented technology

The specification provides a vector graph index merging method and device, the method comprising: obtaining a plurality of vector graph indexes to be merged; wherein each node in the vector graph index represents a vector, and each edge represents that the vectors represented by the nodes connected by the edge are similar; for each target vector in each target vector graph index, determining the vectors similar to the target vector from each vector graph index based on an approximate nearest neighbor search algorithm to form a cross-graph candidate neighbor set of the target vector; based on the cross-graph candidate neighbor set of each vector in the plurality of vector graph indexes, constructing a global neighbor graph; wherein each node in the global neighbor graph represents each vector in the plurality of vector graph indexes, and each edge represents that the vectors represented by the nodes connected by the edge are similar; and converting the global neighbor graph into a merged graph index that can be used for approximate nearest neighbor search.

The application discloses a distribution network fault identification method and system based on transient state feature, and belongs to the technical field of power system automation and intelligent operation and maintenance. The method obtains distribution network transient recording data and extracts multi-dimensional features, and constructs a state description vector mapping the underlying physical discharge state. The state description vector is combined with the multi-dimensional features to construct a mechanism collaborative feature subset, and redundant features in a random discrete state are removed in combination with distribution network grounding operation parameters. The remaining subset is projected into a transient latent variable space, and a state membership degree vector of the subset and a preset physical state anchor point is calculated. The state membership degree vector is converted into a control mask to inject a deep neural network, and a joint loss function is used to constrain the network hidden layer feature mapping result to fit the corresponding physical state anchor point, so that an initial classification is output. Finally, the feature marginal contribution degree of an inference link is calculated, and a final diagnosis result is output. The application realizes deep integration of physical mechanism and artificial intelligence, and solves the problem of lack of physical basis of a black box AI model.

The application provides a vector map construction method and device, a storage medium and electronic equipment, comprising: performing low-light image enhancement processing on N frames of current images to obtain N frames of enhanced images; performing multi-view BEV conversion processing on the N frames of enhanced images to obtain BEV current features; performing fusion processing on the BEV current features and BEV features in a historical window to obtain BEV fusion features; generating a vector curve based on the BEV fusion features, and completing vector map construction based on the vector curve. In a low-light environment and a dynamic scene, the perception ability of an automatic driving system, the stability of map construction and the calculation efficiency can be effectively improved; through fusion of high-definition enhanced images and space-time information, high-precision and low-delay vector map prediction is realized, the robust application requirements of automatic driving technology in a complex environment are met, and more reliable technical support is provided for the development of a future intelligent transportation system.

This application discloses an autonomous navigation and deception defense system and method for unmanned aerial vehicles (UAVs) in GPS-denied environments. The system includes a multi-source sensor array, and parallel deception detection and isolation threads and a tightly coupled multi-source fusion navigation thread. The deception detection and isolation thread performs multi-dimensional consistency checks on received navigation data. When a dual comparison of geomagnetic fingerprints indicates a poor match between GPS position and geomagnetic readings, but a good match between visual position and geomagnetic readings, the GPS data is isolated from the fusion module. The tightly coupled fusion navigation module employs an error state Kalman filter with an IMU as its core. A geomagnetic deviation term is introduced into the state vector, and the original observation geometric information is directly fed into the filter for tight coupling. The difference between the geomagnetic vector obtained from a geomagnetic vector map query and the measured value from the magnetometer is used for fusion and updating, achieving high-precision continuous navigation in GPS-denied environments.

This disclosure provides a method and apparatus for drawing a vector map of an accident scene. The method includes: performing object recognition on a global top-view of a target accident scene to determine at least one object included in the target accident scene; symbolizing the at least one object in the global top-view to obtain at least one object symbol; determining at least one reference element based on the distribution of the at least one object in the global top-view; determining annotation information for at least one key accident point in the global top-view based on the at least one reference element; and obtaining a symbolized vector map of the scene with annotation information based on the annotation information of the at least one key accident point and the at least one object symbol. This disclosure enables the automatic drawing of a standard vector map of a target accident scene while automatically annotating the details of the accident scene.

The application discloses an agent action prediction method based on multi-scale space perception, comprising map space structure modeling, historical trajectory feature extraction, feature fusion and multi-modal action prediction, wherein: the map space structure modeling uses a multi-scale graph convolutional neural network to extract map features from two-dimensional vector map data in an application scenario map, to obtain high-dimensional map feature information; the historical trajectory feature extraction uses a convolutional neural network and a feature pyramid network to extract high-dimensional trajectory data features of all agents; the feature fusion models and fuses the correlation of high-dimensional map feature information and high-dimensional trajectory data feature information through a self-attention mechanism, to obtain agent trajectory fusion features with direction information; and the multi-modal trajectory prediction uses the agent trajectory fusion features for prediction regression and confidence scoring, to output multi-modal complete trajectory coordinates and corresponding confidence scores of action prediction, so as to provide reasonable auxiliary decision-making for agent action.

This application discloses an image annotation method, system, device, and medium, relating to the field of information processing technology. The method includes: acquiring a first set of line features in a lane-level map and a second set of line features in a vector map; converting the first set of line features to a vehicle coordinate system to obtain a third set of line features, and converting the second set of line features to a vehicle coordinate system to obtain a fourth set of line features; matching the third set of line features and the fourth set of line features based on geometric distance to obtain multiple line segment pairs; acquiring at least one corresponding point pair for each line segment pair to obtain a point pair set; obtaining geometric correction parameters that minimize the overall difference of the point pair set based on an optimization model; correcting the third set of line features based on the geometric correction parameters to obtain a fifth set of line features; and projecting the fifth set of line features onto the environmental image to generate annotation data for the environmental image.

The application discloses a kind of based on the method for jumping and fusing positive and negative obstacles segmentation of multi-task guidance, including the following steps: build the jumping and fusing positive and negative obstacles segmentation model based on multi-task guidance and training;Surface normal vector map is calculated by parallax map / depth map, after resolution adjustment to surface normal vector map and corresponding RGB image, input to the jumping and fusing positive and negative obstacles segmentation model trained, obtain positive and negative obstacles segmentation result;The jumping and fusing positive and negative obstacles segmentation model adopts normal vector branch to receive surface normal vector map, and output normal vector segmentation map is used for loss calculation in training process;RGB branch is received using RGB branch, and output RGB segmentation map is used as output positive and negative obstacles segmentation result, and RGB segmentation map participates in loss calculation in training process.The application solves the problem that multi-modal network is not good in the scene of inconsistent data between modes Fusion effect, improve the model recognition ability to obstacle.

This invention discloses a robot navigation map generation method based on the fusion of BIM cloud and point cloud, comprising: preprocessing the original BIM model in the cloud; configuring slicing parameters, extracting cross-sectional contours to obtain the original point sequence; constructing a composite threshold model, recursively simplifying the original point sequence, and generating a static vector map after boundary optimization; acquiring point cloud data and performing filtering and mask removal, uploading candidate point clouds of non-static mask areas to the cloud; constructing static mask areas, performing background subtraction and clustering on the candidate point clouds, extracting the convex hull contours of the clusters and performing vectorization simplification to generate dynamic vector boundaries; fusing the static vector map and dynamic vector boundaries in the cloud to generate a fused vector map, and incrementally distributing it to the robot terminal as needed; the robot terminal combining the fused vector map to realize global path planning, local obstacle avoidance, and multi-robot collaborative navigation. This invention solves the traditional BIM transmission and storage problems and the multi-robot synchronization problem.

Embodiments of the present application disclose a multi-source vector map fusion method and device. The method of the embodiments of the present application is to obtain a global road skeleton, a historical vehicle trajectory set and original vector maps from multiple data sources, construct a point graph and a line graph of each original vector map respectively, extract vector element features, element spatial features and local topological features from the point graph and the line graph, fuse the vector element features, the element spatial features and the local topological features to obtain a vector feature vector, extract global topological features from the global road skeleton, extract trajectory time sequence features from the historical vehicle trajectory set, match the vector feature vector with the global topological features and the trajectory time sequence features based on a preset feature matching algorithm, and fuse multiple original vector maps with a matching rate higher than a preset matching rate to obtain a target vector map. The method additionally analyzes the global road skeleton and the historical vehicle trajectory set to participate in the fusion of the original vector maps, thereby ensuring topological accuracy and consistency.

This invention discloses a method and system for lane line vectorization pre-annotation based on multi-frame LiDAR point clouds. It constructs a global dense point cloud map by performing spatial coarse filtering and spatiotemporal synchronous stitching on the original point cloud data; extracts the main road surface point cloud through elevation fine filtering and plane segmentation; suppresses and enhances non-uniform backgrounds based on morphological top-hat transform; extracts candidate lane line point clouds using adaptive threshold segmentation combining statistical lower bound and recursive maximum inter-class variance method; extracts structured candidate lane lines and geometric features based on weighted iterative RANSAC line fitting guided by both intensity and longitudinal density; and constructs a global topology graph model and a multi-dimensional scoring mechanism to screen and resolve conflicts among candidate lane lines, outputting the final lane line vector map. This invention effectively overcomes uneven intensity and local noise interference, achieving high-purity and high-robustness lane line vectorization reconstruction, and significantly reducing the cost of manual annotation.

The application relates to the field of electric digital data processing and discloses a sensing data fusion abnormality detection system and method based on edge computing, which comprises a data acquisition module, a feature storage module and a data analysis module. The data analysis module acquires N-path heterogeneous digital signals of the real-time state of a controlled object, constructs an observation vector mapped to an N-dimensional feature space through normalization processing, calls a pre-stored coupling feature matrix, projects the observation vector to a stable manifold space, extracts an orthogonal residual vector of the observation vector and calculates the module length, and when the module length continuously exceeds a judgment threshold for a period reaching a time threshold, it is judged that the controlled object has nonlinear structural decoupling. The application identifies abnormalities by monitoring the topological offset of the observation vector relative to the stable manifold, realizes deep mining of the physical coupling logic among multi-source signals, and effectively resists signal slow drift caused by environmental fluctuations.

The embodiment of the invention provides an unmanned aerial vehicle video geographic information superposition method and device based on deep learning, and relates to the technical field of unmanned aerial vehicle visual enhancement and geographic information fusion. According to the method, after unmanned aerial vehicle detection data are obtained, a current view field model is constructed according to positioning data and flight state data, and vector map data are retrieved from a map database based on the view field model. And performing cross-modal fusion on the visual semantic features of the video frame image and the map structure features of the vector map data by using a space detection model based on deep learning, and establishing space corresponding information between visual contents and geographic elements based on implicit reasoning. And then, according to the spatial corresponding information, superposing the geographic information to the video frame image to generate a superposed frame image. According to the method, pixel-level accurate alignment of the geographic information in the video frame image can be realized by utilizing visual semantics and the cross-modal fusion capability of map priori, and the result deviation of video geographic information superposition is reduced.

This invention relates to the field of building outline extraction technology, and discloses a method, apparatus, electronic device, and storage medium for extracting building roof outlines. The method includes: extracting building boundaries from optical remote sensing images and extracting building vector frames from vector maps; performing a global search within an initial search space using a normalized cross-correlation similarity metric, performing template matching, and initially locating the building roof outline; calculating the global offset principal direction angle, setting an angle tolerance, generating a fan-shaped search space with the global offset principal direction angle as the central axis and the angle tolerance as the half-width; performing local optimization within the fan-shaped search space using a normalized cross-correlation similarity metric, performing template matching, iterating the angle tolerance, converging the fan-shaped search space, and repeating template matching to obtain the final building roof outline. This invention implements hierarchical matching from global search to local optimization, and uses a normalized cross-correlation similarity metric for template matching to achieve building roof outline extraction.

This invention discloses a highway tolling processing method, system, and electronic device, comprising: receiving a tolling request from a road network node tolling terminal relative to a target vehicle, and sending a request to a data platform to obtain the target vehicle's driving information; receiving driving location information fed back by the data platform and road network node information along the route after verification according to request verification rules; reconstructing the target vehicle's driving path based on a pre-constructed road network model vector map, the driving location information, and the road network node information along the route, to obtain the reconstructed driving path; determining the tolling result of the target vehicle based on the reconstructed driving path and a pre-set toll rate benchmark, and feeding it back to the road network node tolling terminal. The above technical solution ensures the accuracy of the target vehicle's location result, improves the accuracy of path reconstruction, and enhances the success rate and accuracy of online tolling.

The application discloses a deicing vehicle chassis intelligent path planning method and system, and realizes automatic driving of the chassis of the deicing vehicle in a deicing process. The method comprises the following steps: step 1, loading a high-precision vector map of an airport; step 2, converting the high-precision vector map into a global road network topology map; step 3, performing path searching based on the global road network topology map; step 4, acquiring sensor data and identifying dynamic obstacles; step 5, performing local tracking and safety evaluation; step 6, performing dynamic obstacle avoidance in a lane; step 7, when the deicing vehicle reaches a target airport apron intersection, the deicing vehicle leaves a fixed service lane and enters a semi-open area around the apron, and is switched to a local cost map; step 8, generating a virtual safety envelope line according to an aircraft model; and step 9, using a local path planning algorithm and combining the virtual safety envelope line to enable the deicing vehicle to perform multi-degree-of-freedom fine adjustment to approach the aircraft body.

This application belongs to the field of land use planning verification technology, and discloses a method, apparatus, storage medium, and equipment for land use planning verification. It involves acquiring a planning vector map and a color block planning map of the area to be verified; inputting the color block planning map into a convolutional neural network to extract color block features, obtaining a multi-scale feature map; inputting the multi-scale feature map into a pre-trained semantic segmentation network for color block segmentation, outputting color block regions; constructing a topological structure map of the area to be verified based on the color block regions; inputting the topological structure map and the planning vector map into a graph neural network to analyze topological relationships, and obtaining a state prediction result for each target node by aggregating the features of the target node itself and the features of its neighboring nodes. The state prediction results include compliance, attribute violation, spatial violation, or combined violation; integrating the state prediction results, and generating a verification report based on a rule engine. This can improve the efficiency of land use planning verification tasks and enhance the consistency and stability of verification results.

This application discloses a vehicle control method, device, electronic device, and storage medium, comprising: acquiring sensor data of the vehicle's environment; generating a bird's-eye view vector map; obtaining a predicted trajectory and driving style vector of the vehicle based on the current bird's-eye view vector map and a pre-determined historical operation sequence of the vehicle; obtaining a predicted trajectory of traffic participants based on the current and historical bird's-eye view vector map sequences; generating a control strategy for the vehicle and a set of control parameter adjustment instructions corresponding to the control strategy; and controlling the vehicle to execute the control strategy and the set of control parameter adjustment instructions corresponding to the control strategy. The embodiments of this application generate a high-precision bird's-eye view vector map, using the structured bird's-eye view vector map as an intermediate representation to accurately predict the intentions of the vehicle's driver and traffic participants, generating a clear and effective control strategy and its control parameters, and achieving intelligent, personalized, and safe collaborative control of the vehicle.

The application discloses a kind of based on the automatic navigation map construction method of unstructured mountainous scene-oriented vehicle-mounted equipment, belong to the field of automatic driving and mobile robot navigation.The vehicle-mounted equipment is connected and obtains vision sensor and laser radar data;Including: image semantic segmentation based on prompt word driving, utilize the pre-training of large-scale visual segmentation model that vehicle-mounted equipment is internally deployed in combination with text prompt word set, zero sample semantic segmentation is carried out to RGB image;Through space-time synchronization and semantic point cloud fusion, two-dimensional semantic label is accurately mapped to three-dimensional laser radar point cloud, and global semantic point cloud map is constructed using SLAM pose;Finally, two-dimensional vector and three-dimensional topology are extracted, road centerline is automatically extracted from semantic point cloud, aggregated building contour is detected cliff area, and the two-dimensional vector map of GeoJSON format and the three-dimensional road topology network containing elevation information are generated.The application realizes the full-process automation construction from sensor data to navigation map in vehicle-mounted end, significantly improves the map construction quality and navigation safety in unstructured scene.

The application relates to the field of electric digital data processing and discloses a sensing data fusion abnormality detection system and method based on edge computing, which comprises a data acquisition module, a feature storage module and a data analysis module. The data analysis module acquires N-path heterogeneous digital signals of the real-time state of a controlled object, constructs an observation vector mapped to an N-dimensional feature space through normalization processing, calls a pre-stored coupling feature matrix, projects the observation vector to a stable manifold space, extracts an orthogonal residual vector of the observation vector and calculates the module length, and when the module length continuously exceeds a judgment threshold for a period reaching a time threshold, it is judged that the controlled object has nonlinear structural decoupling. The application identifies abnormalities by monitoring the topological offset of the observation vector relative to the stable manifold, realizes deep mining of the physical coupling logic among multi-source signals, and effectively resists signal slow drift caused by environmental fluctuations.

This invention provides a method for detecting overlapping regions in 3D SAR, including collecting and processing image area data; identifying overlapping regions in each image within the area data and merging initial overlapping vector maps; using an overlapping homogeneity coefficient to compare and filter images in the planar dimension to obtain a final overlapping vector map; calculating elevation difference maps by combining processed interferometric DSM data products and point cloud DSM data products within the image area data; calculating slope difference maps by combining processed interferometric DSM data products and point cloud DSM data products within the image area data; and comprehensively verifying the final overlapping vector map, elevation difference map, and slope difference map to obtain the final overlapping region. This invention identifies and verifies overlapping regions in both planar and elevation dimensions, uses an overlapping homogeneity coefficient to filter non-overlapping regions in the 3D SAR DOM data products, and proposes a comprehensive weighting method to calculate the elevation difference between interferometric DSM data products and point cloud DSM data products to determine overlapping regions.

This invention relates to the field of urban underground space development technology and proposes a method for evaluating the geological safety of underground space, comprising the following steps: S1, establishing a hierarchical underground space geological safety evaluation index system; S2, obtaining basic geological data vector maps for each secondary index to form a hierarchical vector map; S3, forming a comprehensive vector map, dividing the evaluation units, and obtaining hierarchical data for each secondary index within each evaluation unit; S4, obtaining subjective weights using the G1 method, obtaining objective weights using the improved CRITIC method, and finally obtaining the combined weights of each secondary index; S5, constructing a Bayesian network structure; S6, obtaining the prior probability of the root node and the conditional probabilities of other nodes to complete the construction of the Bayesian network model; S7, calculating the underground space geological safety score for each evaluation unit. This invention avoids the limitations of single weighting and solves the challenge of obtaining basic data when using traditional Bayesian network models to evaluate the geological safety of underground space.

This invention relates to the field of traffic information processing technology, specifically to a traffic generation prediction method based on GIS and image vectorization. The method includes georegistring a color raster image of a planning area to obtain a calibrated image; selecting color regions from this image to construct a training sample set labeled with land use type; training an image classification model using a GIS platform and reclassifying the calibrated image; aggregating adjacent pixels of the same type in the classification results into independent vector plots to form a plot vector map; calculating the actual geometric area of ​​each plot in the map and aggregating them according to traffic zones; and finally, combining the traffic parameters and area of ​​each plot to calculate the future traffic generation and attraction of each traffic zone. This invention solves the problems of traditional methods relying on vector planning files, low efficiency, and poor classification accuracy, and achieves automatic and accurate extraction of land use information and traffic generation prediction from easily accessible planning images.

The application relates to the technical field of image analysis, in particular to a deep learning remote sensing image analysis method and system for automatic extraction of land use change, which comprises the following steps: multi-scale spatial features of double-phase original remote sensing images of a target region are extracted through a convolutional neural network to obtain corresponding deep feature maps; an initial difference feature map is generated through feature vector difference calculation, and a change-sensitive feature tensor is obtained by attention-weighted fusion of the deep feature map; feature enhancement is realized through global average pooling and cross-scale feature integration; edge-guided correction is performed on enhanced feature representation in the up-sampling process according to the edge feature map extracted in the convolutional neural network to obtain a change detection probability map; and a land use change vector map is obtained through contour vectorization of the change detection probability map; and the application can improve the efficiency of deep learning remote sensing image analysis for automatic extraction of land use change.

The application discloses a new energy heavy truck charging optimization management method and system, and belongs to the technical field of electric energy management of new energy vehicles. The method comprises the following steps: in a task initialization stage, high-precision vector map data of a predetermined driving path is acquired, three-dimensional topological feature sequences are extracted, and key nodes are marked; in a parameter acquisition stage, real-time load mass is solved through an airbag pressure sensor, and a current state of charge of a battery is acquired; in an energy consumption modeling stage, the path is divided into micro-segments, a whole-process theoretical energy consumption value is calculated based on a vehicle longitudinal dynamics model, and environmental temperature correction and downhill energy recovery are introduced; in a target setting stage, a target energy supplement state of charge is determined in combination with the whole-process theoretical energy consumption value, the current state of charge and a safety redundancy coefficient; and in a charging execution stage, a charging pile is controlled to accurately charge according to the target value. The application deeply integrates path terrain features and real-time load, guarantees energy safety under complex working conditions, and prolongs the cycle life of the battery.

This disclosure describes a method for estimating the pose of a client device using a magnetic field vector map. The method comprises receiving multiple magnetic field measurements from multiple client devices, each magnetic field measurement representing a magnetic field vector at a geographical location. The method further comprises grouping the magnetic field measurements into one or more region groups, aggregating the magnetic field measurements within each region group to generate a probability distribution of magnetic field vectors associated with the geographical region, determining the magnetic field vectors within each geographical region, and generating a magnetic field vector map. Based on the magnetic field vector map, the method may include estimating the pose of the client device based on the user location of the client device and the magnetic field vectors received from the client device.

This invention discloses a method and system for monitoring surface subsidence in mining areas based on UAV photogrammetry, belonging to the field of photogrammetry technology. The method includes: acquiring geometric and radiometric information based on UAV cruise detection and completing pixel-level alignment to construct a joint observation matrix; relying on a subsidence monitor, conducting curvature change analysis under decomposition mode and deformation component decision based on the joint observation matrix of two monitoring nodes to obtain a curvature flow field vector map; performing real-time monitoring of the mining area and prediction of subsidence precursor risk signals based on this vector map, and completing the subsidence precursor judgment and monitoring and early warning management by calibrating critical streamlines and analyzing their time series anomaly patterns. This invention solves the technical problems of insufficient accuracy in monitoring surface subsidence in mining areas, difficulty in accurately separating interference, and inability to identify subsidence precursors in advance by traditional monitoring methods, achieving high-precision, full-area monitoring and proactive risk early warning of surface subsidence in complex open environments.

This disclosure provides a multi-view stereo matching method based on optical flow consistency, belonging to the field of image processing technology. Specifically, it includes: performing downsampling using an image pyramid algorithm; calculating the depth map and normal vector map of each downsampled image; selecting the best neighboring image; calculating the optical flow motion vector of the downsampled image based on the dense optical flow algorithm and the best neighboring image; constructing optical flow consistency; jointly estimating the optical flow consistency and photometric consistency; recalculating the depth map and normal vector map of each downsampled image according to a cost function; comparing the obtained depth map and normal vector map using a detail restorer, and selecting the depth value and normal vector for each pixel using a selection function to obtain the final depth map and normal vector map for the next layer; finally, fusing the depth map and normal vector map into a three-dimensional point cloud model. This disclosure improves the accuracy and adaptability of image matching.

This invention relates to the field of industrial vision technology, specifically a method for monitoring the quality of high-temperature resistant industrial wire harnesses based on visual analysis. The specific implementation process includes: acquiring a sequence of wire harness images; constructing a response image tensor using optical flow frames; decoupling the Lambertian volume and specular reflection into a reflectivity component map and a surface normal vector map; constructing a cylindrical plane inverse projection transformation matrix based on the wire harness central axis equation; inputting the corrected data into a texture geometry segmentation network to generate a non-defect semantic mask and outputting a wire harness defect feature map; generating topological features of the defect region based on the Hessian matrix for defect monitoring of the wire harness. This invention utilizes photometric stereo decoupling and semantic mask suppression to effectively eliminate high-brightness interference and normal texture noise from the surface of highly reflective wire harnesses, improving the difficulty of defect extraction in low-contrast environments. Through diffusion enhancement and topological descriptors, the sensitivity of wire harness defect recognition is improved, and the false alarm rate of defect detection is reduced.

The application belongs to the technical field of automatic driving, and particularly relates to a vector map registration positioning method and system of a graph attention neural network, and comprises the following steps: acquiring a vector map, wherein the vector map comprises a global vector map and a local vector map; encoding the vector map through a vector map local encoding network and a vector map global encoding network connected in sequence to obtain a vector map local feature and a vector map global feature; finding a mapping relationship between a graph structure feature of the global vector map and a graph structure feature of the local vector map, and obtaining pose information after attention aggregation and attention pose decoding processing. The application can use the vector map to perform a registration positioning task.