569 results about "Landmark" patented technology

Three-dimensional outdoor positioning can be considerably enhanced by using various 3D imaging techniques, LiDAR ranging measurements of surrounding landmarks, and Inertial Measurement Units (IMUs) reaching accuracies in the centimetres range. These technologies, in conjunction with global navigation satellite systems (GNSS) such as GPS, GLONASS, Galileo, and BeiDou, provide autonomous geo positioning. It is crucial to acknowledge the constraints of dead reckoning, which may result in mistakes in positional precision. Integrating modern imaging and correction techniques can significantly improve the accuracy and dependability of outdoor localisation. The amalgamation of various 3D imaging methodologies with LiDAR enhancements improves the comprehension and ranging of the environment, tackling the difficulties presented by GNSS constraints. The integration of IMUs with satellite systems ensures the provision of continuous traveling positioning data, even in instances of obscured or faulty satellite signals. This comprehensive approach improves outdoor location precision and expands the use of these technologies across other domains, such as autonomous navigation, robotics, and geospatial analysis, resulting in more effective and dependable positioning solutions. The hardware system consists of a swarm of sensors, including GNSS, INS, IMU, LiDAR, stereo imaging, and RGB technologies, which are combined to generate three-dimensional colored representations of natural landscapes and urban environments. This system reconstructs topographical features, enabling detailed mapping and analysis. It produces comprehensive three- dimensional sensory data, including geo-location and POI identification, and incorporates advanced image processing techniques, such as edge detection, for precise spatial awareness and environmental understanding. The system is designed with a compact form factor, making it easily integrated into wearable devices and autonomous vehicles. The integration of these technologies forms a sophisticated network of swarm intelligence, crucial for autonomous vehicle positioning and navigation. Particle swarm optimization (PSO) is used to address non-deterministic navigation, with different coupling fusion methods available. The Neuro-fuzzy PSO AI Positioning and Navigation System generates three-dimensional sensory data, providing detailed information about geographical distribution and dead reckoning areas.

A system for computer assisted navigation during surgery includes computer platform that operates to identify a set of locations at which a navigated instrument is palpitating a landmark defined on a surface of a pelvic bone of a patient. Further operations determine a center of rotation for a pelvic acetabulum of the patient based on the identified set of locations at which the navigated instrument is palpitating the landmark. Operations determine an orientation of an anterior pelvic plane (APP) and / or a functional pelvic plane (FPP) of the patient based on the identified set of locations at which the navigated instrument is palpitating the landmark and based on the determined center of rotation for the pelvic acetabulum.

In estimating a position of a vehicle utilizing a global navigation satellite system (GNSS), it is desirable to exclude outliner GNSS measurements due to navigation signals via non-light-of-sight paths from the GNSS to the vehicle, but it leads to a distorted satellite geometry distribution. Complementariness between low-lying visual landmarks and healthy but high-elevation satellite measurements is explored to improve the geometry constraint. Measurements of an inertial measurement unit, low-lying visual landmarks captured by a forward-looking camera onboard the vehicle, and healthy but high-elevation satellite measurements are tightly-coupled integrated via sliding window optimization of system states used in a factor graph. To improve estimation performance, good initial guesses of system states are important. As such, initial guesses of velocity set and position set inside a sliding window are estimated simultaneously based on data of Doppler measurement, double-differenced (DD) pseudorange measurement and DD carrier-phase measurement as obtained in GNSS measurements.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for image processing that involves annotating landmarks on real two-dimensional images. In one aspect, the methods include generating a training set of images of an object for landmark detection. This includes receiving a collection of real images of an object, estimating a pose of the object in each real image in a proper subset of the collection of real images, creating a collection of surrogate images of the object for the training set using the estimated poses and a three-dimensional model of the object.

According to the present invention there is provide a method a method for matching features in an image with landmark representations in a landmark map, the method comprising: providing at least one image; extracting from the image one or more features; providing a landmark map comprising a list of landmark representations, wherein each landmark representation is a representation of a respective physical landmark; creating a plurality of policies, each policy comprising at least one matching decision of the feature and the landmark representation, each matching decision having a stage cost; assigning a policy cost to each of the policies, wherein the policy cost is a function of the stage costs of the matching decisions the policy is comprised of; selecting from the collection of the policies the policy with the desired policy cost. There is further provided an assembly having a processor which is configured to carry out said method.

Autonomous ground vehicles that are outfitted with radar sensors and inertial measurement units accurately determine states of the autonomous ground vehicles, e.g., estimates of the vehicles' positions, orientations, or velocities or accelerations along or about one or more axes, based on data captured by the radar sensors and the inertial measurement units. Where objects are detected in radar scans, the objects are determined to be static (or fixed), or dynamic (or moving), and landmarks representing static objects are identified. Constraints on estimates of states may be calculated based on doppler effects, inertial measurement unit effects, or locations of landmarks, and the states may be determined as solutions to optimization problems based on the data and the calculated constraints. Odometry messages representing the determined states may be generated and stored or utilized for any purpose.

The invention discloses a bionic joint humanoid robot navigation positioning method, device and equipment, and the method comprises the steps: monitoring a landing shock wave signal and reflection echo data of a robot, carrying out the time domain expansion of the shock wave signal to form a pulse characteristic spectrum, extracting multipath propagation components, and carrying out the time difference correlation to generate an environment response value. Screening a direct wave component by using an environmental response value, positioning a sound wave propagation trajectory, and detecting impedance mutation to construct an acoustic matrix. A strong reflection zone is identified through node analysis, a sound pressure gradient field is extracted to determine an obstacle boundary, and a landmark feature bit is determined in combination with an environmental response value. And adjusting a dynamic stride frequency interval based on a landmark feature bit, forming an adaptive detection sequence, and performing sparse processing to construct path configuration. The mechanical vibration component of the shock wave signal is used for piezoelectric conversion to generate recycled electric quantity, the recycled electric quantity is matched with path configuration power to determine a sensing period, intermittent sensing is executed to obtain detection data to construct a detection convergence field, and finally an accurate positioning window is determined and converted into position coordinates.

A forestry vehicle includes a number of different sensors, including LIDAR. A two-stage SLAM technique is used in combination with landmark trees to generate a tree map. This SLAM technique avoids large loop closures by resolving point cloud data into an arrangement of landmarks, reducing the computational working set to support real-time operations. A variety of services and visualizations are supported by the tree map.

A computer-assisted medical system includes a user device wearable by an operator. The user device includes a display device configured to present imagery overlaid in an environment of a manipulator assembly, and a sensor configured to detect one or more landmarks in the environment. The medical system includes a controller configured to execute instructions to perform operations The operations include receiving, from the sensor, position or orientation information for the one or more landmarks in the environment, and directing a manual movement of a portion of the manipulator assembly by causing the display device to present the imagery overlaid in the environment based on the received position or orientation information.

The invention relates to the technical field of machine vision recognition, in particular to a dynamic obstacle avoidance method and system for equipment based on behavior trend prediction, equipment and a medium, and the projection range of the equipment on the ground can be effectively adjusted by introducing an image segmentation expansion scale regression mechanism driven by equipment PLC features and an image segmentation expansion prediction model. The method not only improves the fitting precision of the real boundary, but also ensures the consistency of the landmark covered by the expansion boundary, and provides stronger data support for collision modeling of image-physical fusion. A human body positioning and equipment lower edge convex edge construction method based on image bottom edge center point back projection is introduced to serve as a key link of collision modeling, so that the two-dimensional space relation between equipment and personnel is restored. Through the above method, not only is the diversity of different equipment forms paid attention to, but also the directivity and orientation characteristics of the equipment can be accurately modeled, and the stability and interpretation of collision pre-judgment are remarkably improved.

One embodiment of the present invention sets forth a technique for performing landmark detection. The technique includes applying, via execution of a first machine learning model, a first transformation to a first image depicting a first face to generate a second image. The technique also includes determining, via execution of a second machine learning model, a first set of landmarks on the first face based on the second image. The technique further includes training the first machine learning model based on one or more losses associated with the first set of landmarks to generate a first trained machine learning model.

A processor-implemented method including selecting a first reference landmark from among a first plurality of candidate landmarks related to a parking area of a vehicle in a first frame image corresponding to a first timepoint, determining positioning information of the vehicle by using geometric relationship information of the selected first reference landmark with respect to the vehicle, selecting a second reference landmark from among a second plurality of candidate landmarks in a second frame image corresponding to a second timepoint, the second timepoint being temporally subsequent to the first timepoint, and updating the positioning information using the geometric relationship information of the selected second reference landmark with respect to the vehicle.

The invention provides a multi-level semantic map construction method based on multi-sensor fusion, and the method carries out the construction of an environment grid layer, and comprises the steps: constructing an environment grid map in real time through fusing perception data; scene semantic layer construction: extracting image scene semantic probability distribution by using a deep convolutional network, fusing time sequence observation through Bayesian filtering, and mapping a scene category to a grid unit by using an occupation probability model; constructing an object semantic layer, namely identifying an object by adopting a target detection network in which an information aggregation-distribution mechanism is introduced, extracting an object point cloud, and dynamically updating object semantic attributes of grid units through multi-source observation fusion; and scene atlas generation: constructing a hierarchical scene atlas which takes the marker object as a reference core and comprises a spatial topological relation. According to the method, the dynamic environment adaptability and the multi-modal data fusion precision of semantic mapping are improved, a more visual environment understanding mode is provided for the robot, and the practicability of the semantic map in robot positioning and navigation is improved.

The invention provides a positioning method and system based on multi-sensor fusion and a storage medium. Through strategies of local dynamic adaptation, global static correction and multi-source fusion, the problem of laser SLAM positioning degradation caused by warehouse environment changes is effectively solved. Local sub-graph matching adapts to a dynamic environment, and frequent reconstruction of a global map is avoided. And the dynamic weight distribution ensures the positioning accuracy in different environments. The redundant design of multiple sensors such as a laser radar / a wheel type encoder / a two-dimensional code ensures continuous operation in a complex scene. When laser matching fails, the two-dimensional code landmark provides redundant positioning. And flexible addition of new sensors is supported, and requirements of different storage scenes are met. Through combination of local subgraph matching, global observation constraint, an encoder speedometer, a two-dimensional code landmark and multi-sensor weighted fusion, the scheme not only significantly improves the positioning precision and robustness of the AGV in a dynamic environment, but also reduces the dependence on environmental consistency.

In accordance with an aspect of the present disclosure, there is provided a vehicle position estimation method. The method comprises, obtaining a landmark-based initial position information of a camera by matching a landmark in a high definition map corresponding to a GPS-based initial position information of a vehicle to an image obtained by the camera of the vehicle, and obtaining estimated position information of the vehicle by comparing the image with a landmark in the high definition map corresponding to each of a plurality of candidate position information sampled based on the landmark-based initial position information of the camera.

The invention provides a body navigation method, device and equipment based on brain inspiration structured spatial memory, and relates to the technical field of navigation.The method comprises the steps that a work memory module is used for determining a structured spatial memory module for searching a navigation instruction according to the complexity of the received navigation instruction, and the structured spatial memory module is used for searching the navigation instruction; the structured spatial memory module comprises at least one of a landmark memory module and a cognitive map module; responding to the navigation instruction, and performing hierarchical retrieval by using a structured spatial memory module of the navigation instruction to obtain a candidate coordinate set of the target; the candidate coordinate set comprises a plurality of candidate coordinates; determining an exploration sequence based on the confidence score of each candidate coordinate and the distance between each candidate coordinate and the coordinate of the current position of the body agent; the exploration sequence represents the exploration sequence of the candidate coordinates; and controlling the body agent to execute navigation operation based on the exploration sequence. According to the method, the robust space cognition and self-adaptive navigation capability of the intelligent agent with the body in a complex environment is realized.

A method is provided for delivering location-based information using artificial intelligence. The method includes automatically inferring a location of a user system at a geographic landmark based on location data, and triggering an AI assistant in response to the inferred location. The AI assistant generates information about the geographic landmark, and a graphical indication of the AI-generated information is displayed proximate to a graphical representation of the user on a map interface. Upon user selection of the graphical indication, a chat conversation with the AI assistant is initiated, the conversation including the AI-generated information about the geographic landmark

A computer-assisted medical system includes a user device wearable by an operator. The user device includes a display device configured to present imagery overlaid in an environment of a manipulator assembly, and a sensor configured to detect one or more landmarks in the environment. The medical system includes a controller configured to execute instructions to perform operations The operations include receiving, from the sensor, position or orientation information for the one or more landmarks in the environment, and directing a manual movement of a portion of the manipulator assembly by causing the display device to present the imagery overlaid in the environment based on the received position or orientation information.

The invention belongs to the technical field of artificial intelligence, and relates to an underwater navigation method and system based on panoramic visual perception and language model reasoning. The method comprises the following steps: acquiring a 360-degree panoramic video sequence when an underwater robot runs, pre-processing the 360-degree panoramic video sequence, dividing the 360-degree panoramic video sequence into sub-views, extracting visual features, combining position and action embedding information, and aggregating the sub-views into a space-time trajectory feature flow through a Transform network. An adaptive module is embedded in a language model, and visual track features are injected into a language hidden space by using a cross attention mechanism, so that visual-language information fusion and content reasoning are realized. Meanwhile, semantic scene representation is constructed through three-dimensional scene modeling, illusion landmarks are inhibited by means of a landmark uncertainty evaluation module, and finally reasoning output of a language model is mapped into a structured control instruction. According to the method, autonomous semantic navigation based on content understanding in an underwater environment is realized.

The invention discloses an outdoor unmanned aerial vehicle visual language navigation method based on a multi-modal memory map, and aims to solve the problems of memory discretization, insufficient space connectivity and overhigh calculation load caused by dependence on global mapping in an unstructured environment in the prior art. The method comprises the following steps: extracting semantic features from an aerial view image by using a visual language model, and constructing memory nodes in combination with a local truncated symbol distance field; establishing topological connection containing six-degree-of-freedom constraints according to a visual common-view relationship and spatial trafficability among the nodes, and performing hierarchical organization on semantic landmarks in the nodes to form a multi-modal memory map; a semantic weight is calculated based on a cross attention mechanism, and a three-dimensional semantic navigation potential field is constructed by fusing geometric accessibility probability; and determining a navigation target according to the potential field distribution, executing path search in a topology layer, and performing obstacle avoidance and track generation in a measurement layer. According to the invention, the long-voyage autonomous visual language navigation of the unmanned aerial vehicle is realized under the condition that the computing resources are limited.

Implementations relate to methods, systems, and computer-readable media to automatically perform rigging of physics based three dimensional objects. In some implementations, the method may include obtaining a source three-dimensional (3D) mesh of a face of an avatar, wherein the source 3D mesh includes a first plurality of polygons, generating a second 3D mesh based on the source 3D mesh, wherein a count of polygons in the second 3D mesh is fewer than a count of polygons in the source 3D mesh, determining a two-dimensional (2D) parameterization of the second 3D mesh, determining a target 2D parameterization based on a template 3D mesh and the 2D parameterization of the second 3D mesh such that one or more landmarks identified in the source 3D mesh are aligned with corresponding landmarks on the template 3D mesh, and reconstructing the target 2D parameterization into a target 3D mesh.

The invention discloses a landing positioning method and system based on fusion of vision and inertial navigation. The method comprises the following steps: collecting vision information of a current landing area; the visual information of the front landing area is mapped to an aligned grid of a map coordinate system through homography transformation, correlation matching is conducted on the visual information and a map through a frequency domain FFT algorithm, a matching result of rough matching is obtained, and rough matching is completed; the method comprises the following steps of: mapping an area subjected to rough matching and positioning to a grid aligned with a map coordinate system by utilizing homography transformation, performing spatial domain multi-window division on a larger range of images around the area subjected to rough matching and positioning, selecting a landmark area in each sub-window by applying an interest operator, performing spatial correlation matching on the landmark area and a map by utilizing a spatial correlation method, and performing spatial correlation matching on the landmark area and the map by utilizing a spatial correlation method. Finishing fine matching; inputting all matched observation results into an extended Kalman filter, and carrying out joint estimation on the matched observation results and the observation results of the inertial measurement unit; the method has the advantages that the robustness, the precision and the engineering applicability of an existing navigation positioning method are improved.

The invention discloses an unmanned aerial vehicle navigation method and device based on vision and language, terminal equipment and a storage medium, and relates to the field of unmanned aerial vehicle navigation, and the method comprises the following steps: obtaining images of each view angle of an environment where an unmanned aerial vehicle is located and a natural language navigation instruction; the method comprises the following steps: extracting visual features of an image of each view angle through a pre-trained visual language model, and extracting landmark word features corresponding to each landmark phrase in a natural language navigation instruction through a pre-trained large language model; obtaining a potential landmark candidate item corresponding to each landmark phrase through a cross attention mechanism; determining all target landmarks through a pre-trained large language model; obtaining a feasible path according to all target landmarks; and realizing navigation of the unmanned aerial vehicle based on the feasible path. By implementing the invention, the problem that the navigation of the unmanned aerial vehicle in a small crowd scene or an unknown environment is difficult to realize by the existing unmanned aerial vehicle navigation technology can be solved.

A method for generating an updated locator map of a reference scene is provided. The method may comprise: obtaining a preliminary locator map of a reference scene, the preliminary locator map comprising preliminary frames, each preliminary frame being associated with a set of preliminary data points; capturing a plurality of new data points on the reference scene; determining a pose of a capture device related to capture of the plurality of new data points; creating at least one new frame; selecting one or more target frames from the preliminary frame and the new frame; and generating an updated locator map based on the one or more target frames.

In various examples, one or more deep neural networks (DNNs) are executed to regress on control points of a curve, and the control points may be used to perform a curve fitting operation—e.g., Bezier curve fitting—to identify landmark locations and geometries in an environment. The outputs of the DNN(s) may thus indicate the two-dimensional (2D) image-space and / or three-dimensional (3D) world-space control point locations, and post-processing techniques—such as clustering and temporal smoothing—may be executed to determine landmark locations and poses with precision and in real-time. As a result, reconstructed curves corresponding to the landmarks—e.g., lane line, road boundary line, crosswalk, pole, text, etc.—may be used by a vehicle to perform one or more operations for navigating an environment.

The invention provides a dynamic planning method and system for a low-altitude route of an unmanned aerial vehicle, and relates to the technical field of intelligent planning, and the method comprises the steps: receiving a new delivery point instruction, obtaining a three-dimensional coordinate, carrying out the task integration of a new delivery point and an existing delivery point, defining a geographic range through three objects, namely a communication base station, a wireless tower and a permanent landmark building, and carrying out the task integration. Performing space structure division on the geographic range, and disassembling the geographic range into uniformly distributed three-dimensional voxel grids; and generating an adjustment value according to the coordinate interval of the three-dimensional voxel grid, including the delivery point density, the linear distance with the landmark building and other structural attributes, and adjusting the task integration result by using the adjustment value to obtain the adjusted task integration result of all delivery points. According to the invention, safe, efficient and accurate route planning of a multi-delivery-point dynamic newly-added scene in a complex low-altitude environment is realized.

The invention discloses a ship positioning virtual training system based on a sextant simulator. Comprising a 3D modeling reproduction module, a data setting module for setting virtual positioning practice data, a simulation scene construction module for setting a marine virtual training scene, an interactive training module and a positioning model construction module for acquiring an astronomical positioning model and a landmark positioning model. And the comprehensive evaluation storage module is used for acquiring result data and operation grade scores of each operation training stage, and obtaining a virtual training result of ship navigation positioning according to the constructed astronomical positioning model or landmark positioning model based on the result data of each operation training stage. The method solves the problems that traditional manipulation training generally depends on actual ship training, but has certain limitation and problems, due to the reasons that actual ship training is high in cost and high in safety risk, complex navigation environments cannot be truly represented and the like, diversified and flexible operation training scenes are difficult to provide, and the training efficiency is high. And sailors cannot effectively understand and master ship positioning principles, skills and experiences in various complex navigation environments.

Implementations relate to using deep reinforcement learning to train a model that can be utilized, at each of a plurality of time steps, to determine a corresponding robotic action for completing a robotic task. Implementations additionally or alternatively relate to utilization of such a model in controlling a robot. The robotic action determined at a given time step utilizing such a model can be based on: current sensor data associated with the robot for the given time step, and free-form natural language input provided by a user. The free-form natural language input can direct the robot to accomplish a particular task, optionally with reference to one or more intermediary steps for accomplishing the particular task. For example, the free-form natural language input can direct the robot to navigate to a particular landmark, with reference to one or more intermediary landmarks to be encountered in navigating to the particular landmark.

The invention belongs to the technical field of computer vision and intelligent sensing, particularly relates to a positioning system for monitoring blind area multi-sensor space-time calibration, and aims to solve the problems of monitoring blind areas and track breakage caused by non-uniform space-time reference of multi-source heterogeneous sensors. The system comprises a space-time reference generation unit, distributed sensing nodes, a joint calibration engine, a real-time positioning calculation module and a self-adaptive feedback controller. Through nanosecond time synchronization, space registration based on a static landmark and a mobile calibration carrier and a dynamic re-calibration mechanism of closed loop feedback, continuous smooth trajectory output of cross-device target tracking is realized. The system supports layered architecture and block chain evidence storage, improves expandability and security, and is suitable for smart cities and industrial automation scenes.