Analysis of a Hybrid Recognition Method for Detecting Human Motion Posture
This paper describes a motion posture image recognition algorithm that can be used to detect real-time changes in user speed and motion posture.It extracts key information of the target through the camera and combines inertial sensing data for data crossover.It adopts the feature extraction method of frame difference and calibration calculation,normalizes and regresses the processed data,and finally provides the user's possible future motion direction and posture,and predicts and analyzes the motion results.
NETL
NSTL
万方数据
