24 results about "3d tracking" patented technology

The invention discloses a 3D obstacle subdivision attribute pre-labeling method based on a point cloud and an image, and relates to the technical field of attribute pre-labeling, and the method comprises the steps: S1, 3D multi-target tracking; s2, projecting a 3D tracking result to an image; s3, extracting an ROI (Region of Interest); s4, retrieving a subdivision prediction model according to a large category; s5, ROI region splicing and feature extraction are carried out; and S6, fusing multi-frame results and outputting a final result. According to the 3D obstacle subdivision attribute pre-labeling method based on the point cloud and the image, 3D obstacle subdivision attribute pre-labeling is achieved based on the point cloud and the image data, semantic information of the image is fully mined through an ROI region association method and a deep learning network, pre-labeling of the obstacle subdivision attribute is achieved, and the accuracy of 3D obstacle subdivision attribute pre-labeling is improved. And the accuracy is also ensured, and the workload of subsequent manual labeling is reduced.

The invention discloses a lightweight and efficient unmanned aerial vehicle 3D tracking and identification method based on a millimeter wave radar, and the method comprises the steps: carrying out the feature extraction of a millimeter wave signal related to the motion of an unmanned aerial vehicle, and enhancing the difference between the periodic features and noise of the unmanned aerial vehicle through a spectrum folding technology; a Doppler spectrum sequence is obtained after signal processing, the distance of the unmanned aerial vehicle is measured and tracked in real time, and fluctuation of Doppler spectrum intensity and static background noise level in different ranges is reduced; according to the measured distance of the unmanned aerial vehicle, angle position information of the unmanned aerial vehicle is measured and tracked based on an arrival direction algorithm of the subspace; and constructing a lightweight unmanned aerial vehicle identification network model based on a spiking neural network SNN architecture, and obtaining an unmanned aerial vehicle 3D identification result. According to the scheme, the improved angle estimation algorithm is combined with dynamic programming and the particle filter, real-time 3D unmanned aerial vehicle tracking can be achieved, meanwhile, the calculation complexity is reduced, and the tracking calculation time efficiency is remarkably improved.

The invention relates to a monocular vision dynamic SLAM method based on semantics, and belongs to the technical field of computer vision. The method comprises the following steps: 1) carrying out semantic segmentation on a monocular vision image by adopting a deep learning model, and identifying an unknown motion state region and a pure static state region; 2) performing data enhancement on the original image; (3) extracting a SuperPoint feature point by using a SuperPoint network; 4) matching the SuperPoint feature points of the pure static region, and obtaining feature correspondence through confidence coefficient screening; 5) ensuring the accuracy and robustness of subsequent pose estimation through constant-speed motion model tracking, key frame reference tracking, and SuperGlue and SuperGlue-based 2D-3D tracking and repositioning; and 6) constructing a 3D sparse map of the environment. According to the method, an ORB-SLAM 2 system in a monocular camera mode is improved, and a feature point tracking algorithm is optimized.

The invention discloses a nerve stimulation device based on an endoscopic imaging sensor, which comprises an endoscopic imaging sensing electrode module, a tracking label, an external 3D tracking system and an external navigation system, and is characterized in that the endoscopic imaging sensing electrode module comprises a stimulation electrode and an endoscopic imaging sensor installed on the stimulation electrode; the endoscopic imaging sensor captures grayscale images and spectral data from tissue and transmits the grayscale images and spectral data to the external navigation system. According to the nerve stimulation device based on the endoscopic imaging sensor, the problem that the nerve stimulation electrode is inaccurately placed relative to the target nerve is solved, and by combining real-time imaging and navigation tools, the precision of implanting the nerve stimulation electrode is remarkably improved, discomfort is reduced, and the treatment effect is improved.

A 3-dimensional tracking and navigation simulator system for image guided neuro-endoscopy, said system comprising: a fiducial based physical model of an anatomic region associated with neuro-endoscopy, wherein the physical model is mapped with CT / MRI slices of the anatomy; a virtual model of the anatomic region associated with neuro-endoscopy; at least one surgical instrument operable to perform neuro-endoscopy, said at least one surgical instrument is associated with a 3-D tracking component; a camera unit for tracking a location of the at least one surgical instrument with the 3-D tracking component, relative to a reference unit of the physical model; one or more displays for displaying the location of the at least one surgical instrument in the CT / MRI slices and endoscopic output, a computer work-station operably coupled to the physical model, and the camera unit, said work-station comprising: a localization and tracking module having configured to track the tasks performed by a user with the at least one surgical instrument, and estimate the position and rotation of the at least one surgical instrument in 6 degrees of freedom; a virtual space mapping module configured to map a location to a corresponding anatomic position in the virtual model to give a navigation of the surgical instrument in the virtual model; an evaluation module configured to evaluate the 3D tracking of tasks performed based on a localization metric and a 3D tracking metric, and provide feedback to the user based on the performance using machine learning and AI based algorithms.

A tracking system and a method for 3D tracking of a medical device, equipped of at least one ultrasound sensor, inside an anatomical region of a subject having at least one layer of a first tissue type at least partially surrounding a volume having at least one second tissue type, wherein the properties of ultrasounds propagation in the first and second tissue type are different.

Systems and methods are described, and one system includes a three-dimensional (3D) geometric tracker connected to a multiband (MB) inverse synthetic aperture array radar (ISAR), and a classification / alarm logic. The MB ISAR includes spatially distributed radar transmitters (TXs) and receivers (RXs), a TX / RX allocation logic, and a tomographic (TM) image logic. The TX / RX allocation logic is configured to receive 3D tracking data from the 3D geometric tracker, indicating subject 3D position and 3D orientation and, in response, dynamically allocate TXs and RXs to maintain MB illumination of and maintain MB reception of multiple scatter angles from subjects. The TM image processor is configured to construct TM images from the scatter angles, using 3D tracking data, for input to the classification and alarm logic. Optionally, the TX / RX resource allocation logic is configured to receive situation feedback data, for feedback adjusting of allocation of TXs and RXs.

A 3-dimensional tracking and navigation simulator system for image-guided neuroendoscopy, the system comprising: a reference-point-based physical model of an anatomical region associated with the neuroendoscopy (3 and / or 9), wherein the physical model is mapped onto CT / MRI slice images of the anatomy; a virtual model of the anatomical region associated with the neuroendoscopy; at least one surgical instrument suitable for performing neuroendoscopy, wherein the at least one surgical instrument is connected to a 3D tracking component (4); a camera unit (1) for tracking a position of the at least one surgical instrument with the 3D tracking component in relation to a reference unit (5) of the physical model;one or more displays for showing the position of the at least one surgical instrument in the CT / MRI slices and the endoscopic output (6, 12); a computer workstation (11) functionally connected to the physical model and the camera unit (1), the workstation comprising: a localization and tracking module configured to track the tasks performed by a user with the at least one surgical instrument and an endoscope (7) and to determine the position and rotation of the at least one surgical instrument in 6 degrees of freedom (DOF); a virtual spatial mapping module configured to map a location to a corresponding anatomical position in the virtual model to enable navigation of the surgical instrument in the virtual model;an evaluation module configured to assess the 3D tracking of executed tasks based on a localization metric and a 3D tracking metric, and to provide feedback to the user based on performance using machine learning and AI-based algorithms.

The present invention relates to a tracking system (1) and method for 3D tracking of a medical device comprising at least one ultrasonic sensor within an anatomical region of a subject comprising at least one layer of a first tissue type at least partially surrounding a volume comprising at least one second tissue type, wherein the ultrasonic propagation characteristics in the first and second tissue types are different.

This invention discloses a method for real-time 3D tracking and localization of microscale targets in a bright-field microscopic environment, relating to the fields of image recognition and reconstruction technology. The method includes inputting continuous microscopic images into a target detection model, outputting the target's 2D position, regressing the defocus height and direction from a single frame of local images using a fine-grained ROI-based defocus regression model, directly using the model output to drive a 2D motion platform and a focusing device, fusing the 2D position information of each frame with the corresponding defocus axis information, and incorporating the position information of the previous frame as a priori using a temporal correlation mechanism to achieve continuous correlation of the microscale target in the time dimension, thereby stably reconstructing the target's 3D position and outputting a smooth real-time motion trajectory. This invention achieves real-time 3D localization and stable tracking of microscale targets under single-view bright-field microscopic conditions, ensuring both positioning accuracy and system real-time performance and accuracy.

A device and method are proposed that are directed to utilizing 3D tracking approaches and tracker hardware to determine the position and orientation of work tools in 3D space. This position is then validated against a trained position to provide real time feedback on an assembly process. A “work tool” can, for example, be a hand-held tool that is used to perform work, but is not necessarily a hand-held tool. Example work tools include impact drills, torque wrenches, hand-held tooling, or even a human hand.

In various examples, a technique for multiple object tracking is disclosed that includes generating, using one or more processing units, one or more first encoded image features based on a first image. The technique also includes generating a plurality of first object embeddings based on the first encoded image features, wherein at least one first object embedding of the plurality of first object embeddings corresponds to a different object depicted in the first image. The technique further includes determining, using a first track query of one or more track queries and one or more machine learning operations, a first association between a first track and a first object that corresponds to the at least one first object embedding, wherein the first track corresponds to the first track query. The technique further includes computing, based on the first association, an object trajectory associating the first track with the first object.

There is provided an information processing device, an information processing method, and a recording medium that further improve the accuracy of object tracking. A correction unit corrects, based on a tracking result of an object based on 2D tracking using an optical flow and 3D tracking using a depth, errors of the optical flow and the depth, and an update unit updates the tracking result based on the corrected optical flow and depth. The present disclosure can be applied, for example, to a moving body detection device mounted on an autonomous driving vehicle.

The present invention discloses a 3D obstacle segmentation attribute pre-labeling method based on point cloud and image, which relates to the attribute pre-labeling technical field, including S1, 3D multi-target tracking; S2, 3D tracking result projection to image; S3, ROI area extraction; S4, retrieval of segmentation prediction model according to large category; S5, ROI area splicing and feature extraction; S6, multi-frame result fusion and output of final result. The 3D obstacle segmentation attribute pre-labeling method based on point cloud and image realizes 3D obstacle segmentation attribute pre-labeling based on point cloud and image data. Since the method fully mines the semantic information of the image through the ROI area association method and the deep learning network, the obstacle segmentation attribute pre-labeling is realized, and the accuracy is also guaranteed, which reduces the workload of subsequent manual labeling.

A three-dimensional tracking and navigation simulator system for image-guided neuroendoscopy, the system comprising: a standards-based physical model of an anatomical region associated with neuroendoscopy, the standards-based physical model being mapped with CT / MRI slices of the anatomy; a virtual model of the anatomical region associated with neuroendoscopy (3 and / or 9); at least one surgical instrument operable to perform neuroendoscopy, the at least one surgical instrument being associated with a 3D tracking component (4); a camera unit (1) for tracking a position of the at least one surgical instrument with the 3D tracking component (4) relative to a reference unit (5) of the physical model; and one or more displays (11, 12) for displaying the position of the at least one surgical instrument in the CT / MRI slices and the endoscope output. 6, 12), and a computer workstation (11) operatively coupled to the physical model and the camera unit (1), the computer workstation comprising: a localization and tracking module configured to track tasks performed by a user with at least one surgical instrument and to estimate a position and rotation of the at least one surgical instrument in six degrees of freedom (DOF), a virtual space mapping module configured to map positions to corresponding anatomical locations in the virtual model to provide navigation of the surgical instrument in the virtual model, and an evaluation module configured to evaluate the 3D tracking of the performed tasks based on localization metrics and 3D tracking metrics and provide feedback to the user based on performance using machine learning and AI based algorithms.

The utility model discloses a simulation education all-in-one machine for virtual reality technology, relates to education all-in-one machine technical field, and comprises a rotating shaft, a base plate and a cleaning brush, the base is installed above the base plate, one side of the base is provided with a seat, the top end of the installation block is provided with a protection shell, the inner side of the protection shell is provided with an all-in-one machine body, and the all-in-one machine body is provided with an air inlet. A groove is formed in the lower portion of the all-in-one machine body, 3D tracking glasses are placed in the groove in the lower portion of the all-in-one machine body, an infrared pen is placed on one side of the 3D tracking glasses, a sliding plate is movably installed at one end of the all-in-one machine body, and a cleaning brush is installed on the inner wall of the sliding plate. Meanwhile, a sliding plate drives a cleaning brush to move left and right on the all-in-one machine body to clean the surface of the all-in-one machine body, meanwhile, a fan works, cleaned dust enters a main pipe through a branch pipe and then enters a collecting box to be collected, and the problem that dust cleaning is inconvenient is solved.

The invention relates to a tracking system (1) and a method for 3D tracking of a medical device equipped with at least one ultrasound sensor within an anatomical region of a subject, the anatomical region of the subject comprising at least one layer of a first tissue type, the at least one layer of the first tissue type at least partially encloses a volume comprising at least one second tissue type wherein the propagation characteristics of the ultrasound waves in the first tissue type and the second tissue type differ.

This document describes techniques and systems that enable grid-based enrollment for face authentication. The techniques and systems include overlaying a three-dimensional (3D) tracking window over a preview image of the user's face displayed via a display device. The 3D tracking window includes a plurality of segments, which persist to correspond to an approximate direction that the user's face is facing. Based on the tracking, segments are highlighted to indicate the approximate direction that the user's face is facing, a camera captures enrollment images of the user's face facing that direction, and embeddings are generated based on the enrollment images and stored in a fixed grid of pose cells corresponding to various facial poses for use in face authentication. Responsive to generation and storage of the embeddings, an indication that the one or more segments are completed is provided.

A controller (210) for tracking an interventional medical device (252) in three dimensions includes a memory (212) that stores instructions, and a processor (211) that executes the instructions. When executed by the processor (211), the instructions cause the controller (210) to execute a process. The process includes determining (S320 / S420), based on an elevation plane in an ultrasound X-plane mode, a first two-dimensional location of the interventional medical device (252) in the elevation plane. The process also includes determining (S320 / S422), based on an azimuthal plane in the ultrasound X-plane mode, a second two-dimensional location of the interventional medical device (252) in the azimuthal plane. The process moreover includes determining (S330 / S430), based on the first two-dimensional location and the second two-dimensional location, a three-dimensional location of the interventional medical device (252). Finally, the process includes modifying (S340 / S440) ultrasound beam patterns fired in the ultrasound X-plane mode based on the three-dimensional location of the interventional medical device (252).

The invention relates to a position monitoring device comprising: a position marker unit comprising a position marker which is at least partially made of a fluorescent position marker material, an excitation unit which is optically connected to the position marker to excite the position marker to fluoresce, a magnetic field unit which is configured to generate a magnetic field, a detection unit which is optically connected to the position marker and which is configured to detect fluorescence radiation emitted by the position marker, a position processing unit which is connected to the detection unit for data exchange, which is connected to the excitation unit for data exchange, and which is connected to the magnetic field unit for data exchange, wherein the position processing unit is configured to derive data from the excitation unit,The magnetic field unit and the detection unit determine the position of the position marker in relation to the magnetic field.

An energy-harvesting compute grid includes computing assemblies that cooperate with mobile energy harvesters configured to be deployed on a body of water. The plurality of energy harvesters are positioned on and move adjacent to an upper surface of a body of water, and the locations of the energy harvesters can be monitored and controlled. The wide-spread gathering by the harvesters of environmental data within that geospatial area permits the forecasting of environmental factors, the discovery of advantageous energy-harvesting opportunities, the observation and tracking of hazardous objects and conditions, the efficient distribution of data and / or tasks to and between the harvesters included in the compute grid, the efficient execution of logistical operations to support, upgrade, maintain, and repair the cluster, and the opportunity to execute data-gathering across an area much larger than that afforded by an individual harvester (e.g., radio astronomy, 3D tracking of and recording of the communication patterns of marine mammals, etc.). The computational tasks can be shared and distributed among a compute grid implemented in part by a collection of individual floating self-propelled energy harvesters thereby providing many benefits related to cost and efficiency that are unavailable to relatively isolated energy harvesters, and likewise unavailable to terrestrial compute grids of the prior art.

A method for reconstructing 3D shape of a longitudinal device using an optical fiber with optical shape sensing (OSS) properties, e.g. Bragg gratings. By attaching the optical fiber to the longitudinal device, such that the optical fiber follows its 3D shape upon bending, known OSS techniques can be applied to reconstruct 3D shape of the optical fiber, and thus also the longitudinal device, e.g. a medical catheter. E.g. the optical fiber, e.g. placed in a guide wire, can be inserted in a lumen of the longitudinal device. Hereby, one OSS system can be used for 3D tracking a plurality of non-shape sensed catheters or other longitudinal devices. In case the longitudinal device is longer than the optical fiber, the position and shape of the remaining part of the longitudinal device may be estimated and visualized to a user, e.g. based on a known length of the longitudinal device, and based on an orientation of an end point of the optical fiber, e.g. using knowledge about the stiffness or other properties of the longitudinal device.

The invention relates to the technical field of recognition and positioning, and particularly discloses a multi-target recognition and positioning method and system for a robot, and the method comprises the steps: firstly converting discrete three-dimensional point cloud data into a dense depth map aligned with a two-dimensional video frame pixel level in real time through a camera and radar equipment; based on this, the image is dynamically segmented into a plurality of depth regions, and a customized slice and scale transformation strategy is generated for each region. And further, far and near targets with great size differences in the original image are respectively converted into image slices with proper sizes for processing. And finally, unified detection and result inversion are performed on the scale-normalized slices, and three-dimensional tracking is performed in combination with the pose of the robot, so that robust recognition and accurate positioning of the multi-scale target are realized, visual scale dramatic change caused by the change of the distance of the target can be actively eliminated, and the positioning accuracy of the target is improved. Therefore, robust recognition and accurate positioning of the multi-scale target are realized.

This invention relates to a 3D tracking method for sea ice instances based on promptable VGGT (Virtual Geometry-Groograph) technology. The method includes: acquiring a sequence of polar sea ice images; performing 3D geometric encoding on the polar sea ice image sequence to obtain a 3D structure embedding vector for the scene; acquiring prompt text; performing language encoding on the prompt text to obtain a prompt semantic embedding vector; inputting the 3D structure embedding vector and the prompt semantic embedding vector into a cross-attention calculation module to constrain the 3D feature information under the prompt semantics; performing 3D geometric decoding on the 3D feature information under semantic constraints; and combining this with a temporal correlation module to achieve cross-time-sequence instance consistency modeling, generating a segmentation mask and instance ID; and performing instance-level fusion of the point cloud based on the segmentation mask to finally obtain the 3D tracking information of the sea ice. This invention can flexibly specify the tracking object and achieve accurate tracking of sea ice of different scales and morphologies.