1141 results about "Odometer" patented technology

The invention discloses an end-to-end monocular visual odometer method fusing space-time semantic information. According to the method, continuous image sequence frames are collected through a color monocular camera, and a multi-information fusion end-to-end deep learning framework is constructed; a heterogeneous training domain is adopted to set various data set course sharing parameter fusion training, continuous image sequences are input, and the end-to-end deep learning framework is dynamically coupled with hidden state feature vectors output historically, so that a feature mapping relation of time sequence perception is formed; and interpretable feature decoupling of the static background elements and the dynamic entity objects in the scene is realized. After iterative feature fusion, the system outputs sparse depth and camera motion poses which conform to scene geometric constraints, so that a camera trajectory estimation model with high robustness and strong generalization ability in a complex environment is constructed. According to the method, the positioning precision and stability of the monocular vision odometer are remarkably improved.

The invention discloses an underwater vision-inertia-acoustic odometer positioning method fused with a DVL, and belongs to the technical field of robot navigation positioning. According to the method, firstly, an underwater image acquired by a camera and carrier motion data acquired by an IMU and a DVL are used as input information to carry out time alignment and space coordinate system conversion; performing feature point extraction on the image to construct a visual re-projection residual error, and dynamically adjusting a weight ratio of visual information based on the number of feature points and the re-projection residual error; calculating IMU data increment between adjacent frames based on pre-integration to construct an inertial residual error; calculating the displacement change between image frames according to the speed information measured by the DVL to construct a DVL residual error; establishing a nonlinear least square optimization function based on Bayesian maximum posteriori estimation; vision, inertia and acoustic speed measurement data are fused at the rear end through a sliding window mechanism, a multi-constraint joint optimization model is constructed, and the robustness and accuracy of a vision-inertia-acoustic odometer system in the underwater environment are improved.

The invention provides an orchard agricultural robot navigation method, and belongs to the technical field of agricultural robots. The method comprises the following steps: constructing a priori map of an orchard; carrying out odometer initialization and rear end correction on the robot; carrying out ground segmentation and obstacle identification; constructing a cost map: constructing the cost map through static obstacles of a prior map and dynamic obstacles after ground segmentation, and realizing visualization of the obstacles; and path planning: assigning an initial point and an end point on the cost map, firstly realizing global path planning, bypassing an obstacle by using local path planning to realize an obstacle avoidance function when a dynamic obstacle appears in the advancing process, and performing real-time trajectory optimization and smooth control. According to the method, accurate positioning and autonomous navigation of the robot in a complex orchard environment can be realized; meanwhile, through the combination of global path planning and local path planning, the robot can flexibly process static and dynamic obstacles, and the robustness of path planning and the effectiveness of a navigation task are ensured.

The invention discloses a fault detection method and system for a mine unmanned vehicle, and relates to the field of positioning navigation, and the method comprises the steps: collecting the data of each sensor; according to the laser radar data, a slam point cloud map is constructed through a cartographic algorithm; according to the IMU data and the vehicle speed data, performing fusion through a Kalman filtering algorithm to obtain odometer data; the preprocessed laser radar data, IMU data, vehicle speed data, camera data and odometer data are converted to a slam point cloud map coordinate system through coordinate conversion, fusion is carried out through a Kalman filtering algorithm, and state estimation of the unmanned vehicle is obtained; taking the state estimation of the unmanned vehicle as a predicted value, comparing the predicted value with laser radar data, IMU data, vehicle speed data and camera data, and calculating an observation residual error; and performing unmanned vehicle fault detection according to the observation residual error. Aiming at the low positioning precision of the mine unmanned vehicle in a complex environment in the prior art, the positioning precision is improved through multi-sensor fusion, fault-tolerant switching and the like.

The invention discloses a multi-modal information fusion odometer construction method for star catalogue positioning, and relates to the technical field of star catalogue patroller positioning. The method comprises the following steps: carrying out space joint calibration on a monocular camera, a laser radar and an inertial measurement unit, reconstructing a laser radar point cloud by using a timestamp of a camera image, and realizing time synchronization of the camera image and the laser radar point cloud; establishing an IMU pre-integration error model; and performing motion compensation distortion removal on the laser point cloud by using an IMU pre-integration result, and extracting geometric features of the distorted laser point cloud based on a neighbor region smoothness calculation method of a fixed measurement distance. By researching a multimodal information fusion odometer method, the accumulative error of motion measurement is reduced, the positioning precision and stability are improved, technical support is provided for design and development of a star catalogue navigation system, and the problems that a single-modal star catalogue positioning method is weak in environment adaptive capacity and poor in algorithm generalization are solved.

The invention discloses a monocular vision inertial positioning method based on fine pre-integration and adaptive vision inertial weight, which comprises the following steps of: firstly, constructing key frames by using parallax by adopting a sliding window method, and processing the output of an inertial measurement unit between adjacent key frames by using fine pre-integration; initializing a visual inertial odometer to obtain external parameters and internal parameters of an initial sensor and scale and speed information of the odometer; normalizing the acceleration and the angular velocity of each pre-integration quantity by adopting an L2 norm, and weighting the components of the covariance matrix constrained by each fine pre-integration by taking the normalized acceleration and angular velocity as adaptive inertia optimization weights; based on the quality of feature matching after optical flow tracking, a visual weight is obtained according to the number of times of feature continuous tracking and changes of gray values before and after the feature continuous tracking, and the size of an information matrix of a visual re-projection residual error is adjusted in a self-adaptive mode; and finally, performing optimization iteration on the camera pose by adopting a sliding window and a factor graph to minimize a residual term based on mahalanobis distance measurement and obtain a final positioning result.

The invention discloses a multi-sensor fusion SLAM (Simultaneous Localization and Mapping) method in a dynamic scene, which belongs to the field of robot perception and comprises the following steps of: preprocessing visual, laser and IMU (Inertial Measurement Unit) data, including time synchronization and space synchronization; a sensor mathematical model is established, a visual inertia subsystem and a laser inertia subsystem are constructed at the same time, and the two systems are fused in a tight coupling mode. The laser inertial subsystem assists the visual inertial subsystem to complete scale recovery, and the visual inertial subsystem assists the laser inertial subsystem to carry out loopback detection. And finally, establishing a global nonlinear optimized factor graph model in a graph optimization mode, and adding a laser odometer factor, an IMU pre-integration factor, a visual factor and a loopback detection factor into a factor graph to obtain global optimal pose estimation and a three-dimensional map. In a dynamic environment, information of a plurality of sensors is fused, so that the influence of a dynamic object on positioning is reduced, and higher positioning precision and robustness are achieved.

The invention relates to the technical field of digital surveying and mapping of constructional engineering and construction quality control, and discloses a house surveying and mapping system for constructional engineering, which comprises a data acquisition module for synchronously generating a panoramic image signal and an inertial measurement unit signal; the preprocessing module receives a panoramic image signal and an inertial measurement unit signal, and fuses the panoramic image signal and the inertial measurement unit signal through a visual inertial odometer to generate a pose calibration signal; the fusion modeling module receives the purification point cloud signal and the pose calibration signal, and generates building component parameterized data based on a deep learning component recognition algorithm; the result output module receives the building component parameterized data and generates a construction deviation analysis report based on a preset tolerance threshold value; the intelligent control module generates an acquisition quality feedback signal through a real-time quality evaluation algorithm to the data acquisition module so as to adjust a scanning path, and sends a dynamic sampling rate control signal to the preprocessing module so as to optimize a calculation load. The problems of low house surveying and mapping efficiency, poor modeling precision and weak field real-time feedback are solved.

The invention discloses a monocular vision inertial odometer positioning method based on self-adaptive mixed vision residual errors, and belongs to the technical field of robot navigation positioning. The method comprises the following steps: firstly, respectively carrying out feature point extraction and pre-integration on data of a monocular camera and an inertial measurement unit (IMU), and selecting a strong pixel gradient region according to a feature extraction result; predicting a candidate feature point search window based on IMU (Inertial Measurement Unit) motion information, accelerating feature matching and reducing mismatching, and selecting a key frame at the same time; introducing a direct method model on the basis of feature extraction and matching to construct a luminosity error; a feature reliability evaluation method is designed, and a dynamic factor is adopted to adaptively select and optimize a hybrid visual residual projection error or a luminosity error in real time; and at the back-end optimization stage, a sliding window-based graph optimization method is adopted to optimize and combine with the optimization model to perform pose estimation, so that the purpose of improving the positioning accuracy of the visual inertial odometer in a low-light weak-texture scene is achieved.

The invention relates to the field of robots, and discloses a mapping method based on multi-sensor fusion, electronic equipment and a storage medium. According to the method, firstly, a laser radar, a millimeter-wave radar and a visual and motion inertial odometer are generated by collecting multi-sensor data, and it is ensured that other sensors can still provide effective information when part of the sensors are degraded due to insufficient illumination, noise interference or extreme dynamic scenes. And secondly, through real-time health degree evaluation, unstable data generated due to sensor degradation or abnormity is automatically eliminated, and it is ensured that only health data is used for generating a pose map, so that the robustness of data fusion is improved. And finally, a factor graph is constructed by utilizing head IMU data, global pose optimization is realized by virtue of short-time high-frequency stability and continuity of the factor graph, and even if other sensors are partially degraded, the accuracy and consistency of overall positioning can be maintained.

The invention provides a binocular vision inertial odometer method based on a direct method in a dynamic environment, and the method comprises the steps: obtaining a prior segmentation mask of a potential dynamic object through a Mask R-CNN instance segmentation network, and judging the real dynamics of the object through a sparse optical flow method. According to self-adaptive dynamic thresholds of front and back frame optical flow changes and included angles of a potential dynamic object, features of a potential dynamic area are distinguished, features extracted on the dynamic object are removed, and residual static feature points are tracked to estimate an initial pose; in direct method coarse tracking, binocular vision information and IMU information are tightly coupled, and a residual equation of luminosity and IMU is jointly created; and combining IMU error information in a rear-end sliding window, constructing a luminosity residual error by adding a method of projecting feature points of a left-view view of a dominant frame to a right-view view of a target frame, and carrying out iterative updating optimization on luminosity parameters and pose changes.

The invention provides an image acquisition method for three-dimensional point cloud reconstruction and component identification driving of a contact network, and particularly relates to the field of image acquisition of the contact network, and the method comprises the steps: constructing a three-dimensional point cloud model of a contact network structure through a two-dimensional laser radar and a high-precision odometer which are installed on a pushing device; identifying a work support, a carrier cable, a dropper, an electric connection and a positioning point through clustering and principal component analysis, and obtaining key parameters such as a guide height, a pull-out value and a support column limit; and in combination with camera view field parameters, a geometric model is established, a pitch angle and a yaw angle are solved simultaneously, and the camera is controlled to complete image acquisition at the optimal angle and position. Through the method provided by the invention, self-adaptive shooting control under structure identification driving can be realized, complete and clear imaging of key components is ensured, blind shooting and repeated shooting are avoided, and the method has the advantages of high precision, high efficiency and strong real-time performance, and is suitable for a complex overhead line system inspection scene.

The invention discloses a multi-source tight coupling SLAM method based on global information in a dynamic scene, and belongs to the field of mobile robots. According to the method, efficient positioning and mapping are realized through multi-source sensor data fusion and algorithm optimization. The method comprises the following steps of: firstly, eliminating invalid point cloud and data distortion by combining a binocular camera, a single-line laser radar, an IMU (Inertial Measurement Unit) and odometer data and utilizing point cloud filtering, dynamic target detection and motion distortion correction technologies, and fusing observation data of the camera and the laser radar; secondly, IMU and odometer data are fused through unscented Kalman filtering (UKF), and high-precision initial pose estimation is output; and finally, improving a Gmapping algorithm, optimizing a grid map updating module by adopting a subgraph fusion and breadth-first search strategy, and enhancing the environmental feature retention capability of the SLAM module. In a dynamic and complex scene, the method has better robustness, the constructed map better conforms to a real environment, and meanwhile, the pseudo obstacle characteristics in the map can be effectively reduced.

The invention discloses an unmanned aerial vehicle autonomous navigation system based on multi-source vision assistance, and relates to the technical field of unmanned aerial vehicle autonomous navigation, and the system comprises the following modules: a terrain recognition module. A camera and a laser radar are used for synchronously collecting environment data, a visual odometer and a three-dimensional point cloud analysis technology are combined, a dynamically updated grating map is constructed, obstacle and topographic features are recognized through point cloud clustering, an improved Dijkstra algorithm is combined with an information entropy evaluation mechanism to plan a global optimal path, a B spline curve is used for eliminating path mutation, and the optimal path is obtained. The method comprises the following steps that: a dynamic speed decision-making module adjusts a flight parameter in real time on the basis of an obstacle distance-response model, and finally realizes trajectory tracking and dynamic deviation correction through closed-loop PID control and visual pose correction, so that full-link closed-loop control of'perception-planning-execution 'is formed, and the method has high precision, strong robustness and real-time response capability.

The invention relates to the technical field of high-precision positioning, in particular to an outdoor immersive text and travel hybrid positioning and dynamic rendering optimization method, and the method comprises the steps: obtaining a positioning signal of a mobile terminal in a target service area through a hybrid positioning model; the positioning signals comprise an ultra-wideband ranging signal, a visual inertial odometer signal, a real-time carrier phase difference signal and an inertial navigation signal; processing the positioning signal through a multi-layer signal processing filtering model to obtain coordinate data, and generating a space coordinate; constructing a user movement situation prediction model based on the space coordinates, generating a navigation prediction track and navigation confidence, and generating a preloading strategy; acquiring geographical environment parameters of the target service area, and integrating the space coordinates, the geographical environment parameters, the preloading strategy and the virtual object through a multi-dimensional virtual-real fusion model to generate an augmented reality space scene; and analyzing the user movement track data and the augmented reality space scene through a multi-modal driving model to generate interactive content.

The invention discloses a path planning and trajectory optimization method for a mobile robot in a non-flat terrain, and the method comprises the following steps: S1, the mobile robot employs a laser radar to detect the surrounding environment, and an SLAM module constructs a three-dimensional map according to point cloud data and odometer data collected by the laser radar in real time; and S2, a trafficability analysis module performs terrain trafficability evaluation based on the generated three-dimensional map, and generates a trafficability layer accordingly. And S3, the global planning module performs path search in the trafficability layer to obtain an initial path of the mobile robot from the starting point to the target point. And S4, a trajectory optimization module optimizes the initial path to ensure that the obtained trajectory meets the kinematics constraint of the mobile robot. And S5, vertically mapping the optimized trajectory to the non-flat terrain surface to obtain a final path for direct tracking of the mobile robot. According to the method, it can be ensured that the mobile robot generates a feasible track conforming to kinematics constraints in the non-flat terrain.

The invention discloses a rapid measurement method and system for geometric irregularity of a track in a turnout zone. The method comprises the steps that complete three-dimensional point cloud data of the surface of a railway turnout steel rail are obtained in real time through a multi-angle line structure laser scanner on a special detection vehicle body; accurate attitude and position data of the detected vehicle body in the measurement process are obtained in real time through an inertial navigation measurement unit and an odometer; obtaining a high-precision pose of a detected vehicle body through a robust dead reckoning method of an IMU (Inertial Measurement Unit) assisted by a speedometer dynamic lever arm value based on a variable wheel-rail contact point; constructing a turnout area three-dimensional point cloud model; extracting geometric irregularity key points of turnout steel rail point cloud from the turnout area three-dimensional point cloud model; calculating parameters of track gauge, horizontal and rail direction irregularity and unevenness; the special detection vehicle body can quickly and stably pass through a turnout area to complete continuous and accurate measurement of geometric irregularity of a track.

The invention relates to a three-dimensional reconstruction method based on a visual odometer and deep learning. The method comprises the following steps: obtaining related parameters of a monocular camera through camera calibration; shooting continuous image frames of the target area under the plurality of camera poses through the monocular camera to obtain image frames of the target area after distortion correction under the plurality of camera poses; estimating relative poses of the camera among different image frames by using a visual odometer according to related parameters of the monocular camera; constructing a cost body of the image frame by using a cost body generation model according to the relative pose of the camera; optimizing the cost body generation model by using the geometric consistency loss and the depth consistency loss, and generating a depth map of the image frame through the optimized cost body generation model; performing noise elimination and hole filling on the depth map of the image frame to obtain an optimized three-dimensional map; according to the method, a stable reconstruction effect can be kept in a complex environment, a low-texture region and a scene with a large visual angle span, and compared with an existing single deep learning reconstruction method, the method has higher robustness and reconstruction precision.

The invention relates to a multi-sensor fusion SLAM method based on high-precision map assistance, and belongs to the technical field of robot autonomous navigation. The method comprises the following steps: acquiring laser radar data, and constructing a multi-level geometric constraint registration algorithm for the laser radar data to obtain a laser radar odometer; obtaining IMU data, constructing an IMU multi-error self-sensing prediction network for the IMU data, and compensating the IMU data according to prediction errors to obtain an IMU odometer; acquiring a high-precision map and camera data, and constructing a registration method based on manifold learning mapping assisted by the high-precision map for the high-precision map and the camera data to obtain a visual odometer based on the high-precision map; according to the adaptive fusion method based on space-time correlation and local optimization, the contribution degree of each sensor in data fusion is dynamically adjusted, and a multi-sensor fusion correction value is obtained in combination with historical stability, trend change and space-time correlation. According to the invention, the problem of low mapping and positioning precision of a single sensor can be solved.

The invention provides an AGV natural scene positioning method and system based on a 3D laser radar, and relates to the technical field of automatic guided vehicle control. The method comprises the following steps: constructing matching sub-graph information, and providing a key reference for subsequent positioning; a visual sensor is used for capturing multiple image data, IMU data and odometer data of the current pose of the AGV are collected in real time at the same time, and the motion state of the AGV is accurately analyzed according to the IMU data and the odometer data. A 3D laser radar obtains scene point cloud data to construct a local map, the local map is registered with a matched sub-graph, initial pose estimation is obtained, finally, the initial pose, a motion state and image data are integrated, final pose information of the AGV is determined through complex operation, multi-source data cooperatively exerts force, and the positioning accuracy and reliability of the AGV in a natural scene are comprehensively improved.

The invention relates to the technical field of intelligent driving, and provides a vehicle control decision-making method and device based on the perceptual fusion technology and electronic device.The vehicle control decision-making method based on the perceptual fusion technology comprises the steps that first positioning information of a vehicle is obtained through data fusion of at least two complementary sensors, the complementary sensor comprises an inertial navigation system, a visual odometer and a laser radar odometer or odometer; correcting the first positioning information through a roadside assisted positioning technology to obtain second positioning information; state information of a target vehicle is acquired through a cellular Internet of Vehicles communication technology; fusing the environment sensing data collected by the vehicle-mounted sensor and the roadside equipment of the vehicle and the state information of the target vehicle, and constructing a global environment model; and making a control decision based on the second positioning information and the global environment model to realize safe vehicle distance keeping and collision avoidance. The method improves the perception, decision and control capabilities of the vehicle.

Disclosed in the present invention is a LiDAR odometer method using a piecewise linear continuous time trajectory. The present invention comprises the following steps: first, using a time window to divide an initial LiDAR point cloud time sequence so as to obtain LiDAR point cloud sub-sequences under a plurality of time windows, and constructing an initial local point cloud map; and then, on the basis of the initial local point cloud map, optimizing LiDAR poses of the LiDAR point cloud sub-sequences under the time windows to obtain an overall continuous time trajectory and a final local point cloud map. The present invention achieves high-precision motion estimation for LiDARs, has high robustness and real-time performance, is applicable to the fields such as robot navigation, simultaneous localization and mapping, automated driving, and provides a relatively simple and efficient solution that can effectively handle the characteristics of LiDARs as streaming sensors.

The invention discloses a pedestrian inertia / vision / satellite integrated navigation method based on a handheld terminal PDR. The pedestrian inertia / vision / satellite integrated navigation method comprises the following steps: calculating to obtain inertial navigation pose information based on strapdown inertial navigation; when the pedestrian is static, constructing an observation constraint of zero-speed motion for correction, and updating inertial navigation position information based on step length information when a new step is carried out; obtaining visual pose information of the pedestrian based on an inertia / visual fusion odometer, obtaining satellite position information of the pedestrian based on satellite positioning, and carrying out coordinate system alignment and step length coefficient correction factor estimation; a pedestrian multi-source factor graph optimization model with a handheld terminal PDR as a core is constructed, inertial navigation pose information and step length information are used as local state constraints, visual pose information and satellite position information are used as global state constraints, and a target optimization function is constructed to obtain a final pose of a pedestrian. The method is applied to the field of pedestrian navigation, has full-scene robust and continuous positioning capability, and can effectively realize high-precision navigation positioning under long endurance.

The invention discloses a vehicle positioning method and system, and relates to unmanned driving. According to IMU data and laser radar data, a fast-lio algorithm is used to construct a global prior map; constructing a local point cloud map according to laser radar data collected in real time; recognizing and rejecting dynamic point clouds, and carrying out NDT registration on residual static point clouds and the global prior map to obtain a laser odometer pose; extracting feature points of the camera data by adopting a shii-tomasi algorithm; a Lucas-Kanade optical flow method is used to acquire motion vectors of the feature points; calculating the speed of the feature points according to the motion vector, removing the dynamic feature points with the speed exceeding a preset threshold value, and obtaining the pose of the visual odometer; reckoning the position and course of the vehicle through a CTRV model to obtain a track reckoning result; and carrying out multi-source heterogeneous data fusion by adopting a non-destructive Kalman filtering algorithm. Aiming at the defects of dynamic environment interference and non-linear motion trail modeling existing in traditional open air and mines, the positioning stability of the unmanned vehicle is improved.

The invention discloses a multi-source navigation data fusion method for an unmanned loader, which comprises the following steps: data acquisition: acquiring environmental perception data of satellite navigation, an inertial measurement unit IMU, a wheel type odometer and a laser radar and camera in real time; performing space-time synchronization preprocessing, realizing multi-source data time synchronization, unifying environment sensing data to a body coordinate system, eliminating abnormal data, and complementing missing data; carrying out dynamic weight calculation, establishing an error model of each sensor, and adjusting a fusion weight by using an error reciprocal exponential weighting method; layering fusion is carried out, a satellite and an IMU are fused through bottom-layer extended Kalman filtering (EKF), a laser radar and a high-precision map are fused through middle-layer ICP, a middle-layer result and a wheel type odometer are integrated through high-layer federated filtering, and high-precision fusion is achieved; and outputting and optimizing a result, outputting navigation data, performing closed-loop optimization, monitoring the health degree of the sensor, and executing redundancy switching when a fault occurs. The system comprises a corresponding processing unit, and a storage medium stores a program for executing the method.

The invention relates to the technical field of unmanned aerial vehicle intelligent inspection path planning, and discloses an unmanned aerial vehicle intelligent inspection path planning method and system based on visual inspection, and the method comprises the steps: collecting an image, and carrying out the differential superposition processing, and obtaining environment visual data; feature correction is extracted to determine a candidate area, and a template is matched to determine the type and severity of an abnormal event; performing local path planning according to an event generation mode switching instruction; according to the trajectory parameters meeting the minimum turning radius constraint, a safe fly-around path is obtained; the method comprises the following steps of: acquiring a real-time image stream, adjusting a speed parameter to obtain an optimized trajectory, processing the real-time image stream and inertial measurement data generated in the optimized trajectory through a visual inertial odometer technology to obtain environment three-dimensional point cloud data, and performing cyclic verification according to the environment three-dimensional point cloud data to obtain an event processing verification result. According to the invention, hierarchical response and dynamic local path re-planning of abnormal events can be realized.

The invention relates to an automatic guided vehicle (AGV) adaptive path tracking method based on MPC. The method comprises the following steps: S1, establishing a kinematic discretization error model based on the kinematic characteristics of the two-wheel differential AGV, processing a continuous kinematic equation by adopting an Euler discretization method, expressing a dynamic change relationship between a transverse deviation distance and an angle deviation in a state-space equation form, and generating a kinematic discrete state-space model; the method has the advantages that the continuous equation is processed by establishing the kinematics discretization error model and adopting the Euler discretization method, the deviation dynamic relation is expressed through the state space, the calculation process is simplified, the precision is kept, sensor data are fused, the wheel type odometer, IMU and laser data are integrated through the extended Kalman filtering algorithm, and the precision is improved. The real-time position and angle deviation are calculated, the positioning accuracy is improved, a model prediction controller objective function is designed, a weight matrix and boundary constraint are introduced according to constraint conditions, and the effect of obstacle avoidance constraint is combined.

The invention provides a land unmanned equipment semantic preserving large model deployment method and system, and relates to the technical field of navigation control. According to the deployment method, a preliminary control suggestion is directly output through a lightweight fusion model, optimization fusion is carried out through an MPC framework, underlying dynamics and environmental constraints, and seamless connection from semantics to control is achieved; in combination with a VILO odometer and instance segmentation, a global map containing dynamic obstacle semantic information is constructed and updated in real time, and an accurate context is provided for planning and decision making; model pruning, knowledge distillation and edge calculation scheduling are adopted, so that a complex multi-modal semantic model can run in real time on an embedded platform; the decision basis from instruction analysis to control execution is recorded and visualized in the whole process, and the transparency and credibility of the system are improved; on-line re-planning, multi-stage fault detection and switching strategies are integrated, and the robustness of the system in a dynamic environment and an abnormal condition is remarkably improved.

The invention provides an SLAM positioning method and system based on multi-sensor fusion, and the method comprises the steps: firstly receiving data of a plurality of sensors, and obtaining initial data including visual feature data, I MU pre-integration data and UWB distance data through anomaly detection and processing; initializing by using visual feature data and I MU pre-integration data, estimating an initial pose, and predicting a UWB anchor point position as a fixed constraint in combination with UWB distance data; inputting the initial data into a sliding window, outputting drifting-free odometer data, judging the motion state of the SLAM system according to the data, and if the motion state is static, executing zero-speed updating to obtain target pose data; and finally, determining a new key frame based on the parallax or visual feature data quantity of the current frame and the previous key frame. According to the invention, the problems of relatively weak robustness and relatively low positioning accuracy of the system facing extreme scenes are solved; by means of motion state judgment and zero-speed updating, the robustness of the system in long-term operation is improved.

Disclosed in the present invention are an image generation method, system and device for subway tunnel clearance, and a storage medium. The method comprises: a LiDAR collecting three-dimensional laser point cloud data along a railway, and an odometer collecting encoder data along the railway; determining current subway tunnel section point cloud data on the basis of a subway clearance detection project, the three-dimensional laser point cloud data and the encoder data; determining a clearance violation area on the basis of the current subway tunnel section point cloud data and clearance; optimizing intrinsic and extrinsic parameters of a camera on the basis of point cloud information corresponding to the clearance violation area, so as to obtain calibrated intrinsic and extrinsic parameters of the camera; and on the basis of the calibrated intrinsic and extrinsic parameters, generating a camera projection map from the current subway tunnel section point cloud data, and constructing an image of subway tunnel clearance on the basis of the camera projection map. The present invention realizes, on the basis of point cloud data and clearance, track clearance violation detection during a high-speed inspection process, thus facilitating inspection and viewing by means of constructing an image of subway tunnel clearance.