10 results about "Inertial motion capture" patented technology

This invention belongs to the field of deep learning, specifically relating to a posture-induced multimodal gait recognition method aimed at improving the accuracy of gait recognition in complex environments. The method includes constructing a multimodal dataset containing gait video sequences, corresponding depth map sequences, and motion-captured posture sequences by combining a full-body inertial motion capture device, a binocular depth camera, and an RGB camera. Cross-device gait feature alignment preprocessing is performed on the multimodal dataset to obtain spatiotemporally aligned motion-captured postures and video postures. A 3D posture induction module is constructed, using the spatiotemporally aligned motion-captured postures as a reference, and trained under supervised training via graph feature alignment loss to extract induced posture information from the video postures. A multimodal recognition model fusing induced posture information and gait contour features is constructed to achieve gait recognition.

The application discloses a kind of human motion digital twin construction methods based on inertial action capture technology.Firstly, action capture device is worn on human body, and the original sensing data of each joint corresponding sensor is collected, the original sensing data is processed using adaptive extended Kalman filter algorithm, the attitude of each joint corresponding sensor is obtained, and the original sensing data and sensor attitude data are sent to host computer;Host computer constructs human motion digital twin model according to human skeleton;The attitude calibration matrix of each joint is calculated using human posture calibration method, and the real-time sensor attitude data is calibrated to obtain the attitude of each joint of human body, drive human motion digital twin model;While driving the model, the joint parameters calculated are compared with human joint parameter threshold, and the evolution model is dynamically updated.The application can realize that human motion digital twin model is synchronously mapped real human motion, and model is updated in real time dynamically.

This application discloses a motion data redirection method, system, device, and storage medium. The method includes the following steps: acquiring the first skeleton coordinates of the target character, the second skeleton coordinates of the original character, the first driving quaternion of the original character in the current frame, the first root node coordinate information of the original character in the current frame, the first whole-body node coordinate information of the target character in the previous frame, and the fourth whole-body node coordinate information of the original character in the previous frame; determining the temporary second whole-body node coordinate information of the target character in the current frame; determining the third whole-body node coordinate information of the original character in the current frame; determining the node identifier with the smallest coordinate change; determining the motion change of the node corresponding to the node identifier in the target character; and adjusting the first whole-body node coordinate information of the target character in the current frame. This method can be applied to both real-time and non-real-time motion data redirection, and has strong practicality and scalability. This application can be widely used in the field of inertial motion capture technology.

This application relates to the fields of motion capture technology and computer human body model technology, and particularly to a method, device, electronic device, and storage medium for automatic adjustment of a surface model. The method includes: acquiring the posture data of a reference human body at the current moment; adjusting the current posture of the surface model of the reference human body based on the posture data to obtain the target posture of the surface model, and acquiring posture representation data of a model human body using an inertial motion capture system; determining the mapping relationship between the reference human body and the model human body, and adjusting the target posture of the surface model based on the mapping relationship and the posture representation data to obtain a dynamic surface model that changes over time. This solves the problems in related technologies where posture human body model modeling requires manual adjustment, has a low degree of automation, and where automatic adjustment of human body models based on optical motion capture is expensive, complex to operate, and has a narrow range of applications.

A robot learning dataset construction method, a humanoid robot, and a computer-readable storage medium are provided. The method includes: a robot operator uses an inertial motion capturing device to remotely operate a humanoid robot to perform various operation tasks, construct a dataset by collecting real-time working images and real-time joint motion data of the humanoid robot performing various operation tasks under manual guidance, thereby utilizing the flexible combination of the human body posture direct extraction characteristics of the inertial motion capturing device and the flexibility and reliability of executing robot remote operation tasks so as to improve the construction efficiency, dataset validity, and dataset integrity of the task learning dataset, which facilitates further improvement of the generalization and accuracy of the imitation learning of the humanoid robot.

The invention discloses an inertial motion capture method, device and system and a storage medium, and the method comprises the steps: constructing a ToF data simulation model, and obtaining node data measured by an IMU sensor and a ToF sensor in real time; node center data fusion is carried out to obtain node sensing features, and data coding is completed based on Transform Encoder according to the node data; performing dynamic spatial position coding, regarding node positions as functions of motion signals, and performing sine and cosine position coding in the network; adding the dynamic space position codes and the node sensing features and inputting the added codes and the node sensing features into an encoder; after dynamic space position coding and node sensing coding are completed, the fusion feature ZNCI is input into three cascaded LSTM motion estimators, and calculation of the speed, the position and rotation is achieved; carrying out adaptive fusion calculation on the end point motion state according to the three cascaded motion estimators; according to the method, the positioning precision of the tail end of the key node can be improved, and attitude features in nonlinear and slowly-varying actions can be captured more accurately.

The application discloses a kind of robot whole body control method and system based on real-time inertial motion capture and reinforcement learning, belong to robot control field.The system includes the motion capture module, data processing and forward kinematics module, motion redirection and optimization module, general motion strategy control module and real-time control and driving module in turn link, form end-to-end closed loop control flow line by shared memory and high-speed communication protocol.The method is realized high-quality mapping from human posture to robot model by collecting human joint data through inertial motion capture equipment, pose solution, non-uniform scaling and two-stage differential inverse kinematics optimization, and generates stable joint instructions combined with reinforcement learning strategy trained by PPO algorithm.The application solves the problems of lack of feedback loop, joint deadlock and contradiction between realism and stability in the prior art, has the characteristics of strong robustness and high expansibility, and can be widely applied to remote operation, motion reproduction and other scenes.

The invention provides a human body linkage immersive VR sightseeing interaction system and a human body linkage immersive VR sightseeing interaction method, and the system employs a motion perception-intelligent decision-scene response integrated interaction architecture. Comprising a leg motion sensing module, a motion-scene collaborative decision-making module and an immersive VR park scene presentation module worn on the head, wherein the leg motion sensing module realizes millisecond-level data linkage by adopting BLE 5.0 connection and is fixedly mounted at the middle section of crus gastrocnemius muscle; wherein the leg motion sensing module comprises a six-axis inertial motion capturing unit used for collecting walking motion characteristics of the old people; the exercise-scene collaborative decision-making module comprises an old people exercise ability adaptation unit with a built-in dynamic threshold model; and the immersive VR park scene presentation module comprises a multi-sensory immersion unit with a built-in natural sound effect library. According to the invention, the solitary feeling of old people during exercise is relieved, and the exercise willingness is improved; the elderly are guided to actively adjust the exercise intensity, and safety and the exercise effect are both considered; the device adapts to the physical characteristics and capabilities of old people.

The invention relates to a regional muscle load grade prediction method and device based on deep learning. The method comprises the following steps: acquiring original motion data captured by inertial motion; carrying out standardized pretreatment on the sample; driving an individualized biomechanical simulation model by using the preprocessed data, and generating a high-fidelity muscle load grade label data set through muscle force calculation, regional aggregation and grade discretization; training a deep learning model according to the data set; and finally, using the trained model to quickly predict new motion data, and outputting the load level of each muscle area. According to the method, through an integrated process, the contradiction between label scarcity and real-time requirements in non-intrusive muscle load assessment is solved, and efficient and interpretable muscle load assessment is realized.

This invention relates to a multimodal spatiotemporal alignment and interactive reconstruction method for digital twins of intangible cultural heritage skills, belonging to the interdisciplinary field of digital protection of intangible cultural heritage and computer graphics. It simultaneously acquires five modalities: optical and inertial motion capture, hyperspectral imaging, panoramic sound field, multi-view images, and physiological data of inheritors. Temporal alignment is achieved through cubic spline interpolation and Gaussian mixture model expectation-maximization algorithm, and spatial alignment is completed using point cloud registration. A dynamic graph spatiotemporal interaction network is used to generate cross-modal fusion features. Based on this, a three-level digital twin is constructed: geometric, technological, and knowledge-based. The technological twin extracts temporal evolution features based on Transformer, while the knowledge twin expresses causal transmission relationships using process hyperedges. Finally, interactive experiences are provided through dynamic geometric optimization, multimodal immersive rendering, and gesture / controller interaction, and user behavior feedback is used for incremental learning and knowledge graph updates, forming a self-evolving closed loop.