Font Size: a A A

Research On Human Motion Tracking And Interaction Technology In Aerospace Testing Missions

Posted on:2018-04-06Degree:MasterType:Thesis
Country:ChinaCandidate:P LiuFull Text:PDF
GTID:2352330512976695Subject:Computer technology
Abstract/Summary:PDF Full Text Request
Aerospace Test Launch Mission is a high-risk,high-tech,high-investment complex system engineering.The operator needs a higher ability in the testing process.Virtual reality technology can build a realistic space environment for the launch site.In order to improve the ability of the operator to operate correctly at the complex site,the operator can drive the virtual avatar in the virtual environment to complete the task of training.In order to implement an operator-driven avatar,human motion tracking needs to be studied.In order to realize the tracking of human motion,a motion tracking system based on inertial sensor and Kinect is developed.The human motion data are collected by the inertial sensor and Kinect.The human skeleton model is established and the initial attitude of the human body is calibrated.The Sensor coordinate system and Kinect coordinate system are converted to human skeletal coordinate.The system also calculates the coordinates of human joints and achieves real-time tracking.In order to improve the stability of human motion tracking system,the inertial sensor data and Kinect data are collected for data fusion,using Kalman Filter.Aiming at the requirement of reliability of motion data in multi-source data fusion,a data fusion algorithm based on data quality evaluation is proposed.The algorithm evaluates the data quality of the inertial sensor and Kinect,and completes the data fusion work according to the evaluation results.When the Kinect or inertial sensor data errors,the fusion algorithm can be modified on the human motion data to ensure that the normal movement of the tracking system stable operation.The validity of the algorithm are verified by experiments.The human motion planning in space launch mission is studied.An RRT algorithm based on random sampling is adopted.Based on the analysis of the problem that the traditional RRT algorithm has a long search time and large scale,an improved RRT-Connect algorithm is used to plan the human action.The experimental results show that the improved algorithm has better performance.The results of this paper are validated in the simulation environment of Space Test Launch Mission.
Keywords/Search Tags:Human Motion Capture, Inertial Sensor, Kinect, Data Fusion, Action Planning
PDF Full Text Request
Related items