Design of Interactive Behavior Target Tracking Algorithm Based on Mixed Reality
In order to accurately track dynamic interactive behavior targets,research on interactive behavior target tracking algorithms based on mixed reality was carried out.Firstly,create a mixed reality environment suitable for target tracking,establish an interactive behavior target motion model,and obtain the target's position,motion trajectory,and interactive behavior information.Secondly,based on the characteristics of target motion and model requirements,extract interactive behavior target features to provide basic data for target tracking.Finally,design a target tracking algorithm to identify the best matching feature points and achieve tracking by matching target features between consecutive frames.The experiment shows that the average accuracy of static target tracking is±0.03 m,and the speed is 35 Hz.The average accuracy of dynamic target tracking is±0.08 m,with a speed of 28 Hz.The accuracy of gesture recognition reaches 95％.The attitude estimation error is only 3°.
NETL
NSTL
万方数据
