Affiliations: [a] School of Physical Education, Yanching Institute of Technology, Langfang, Hebei, China | [b] Physical Education Department, Shangqiu Institute of Technology, Shangqiu, Henan, China
Correspondence:
[*]
Corresponding author: Shanchang Ren, School of Physical Education, Yanching Institute of Technology, Langfang, Hebei 065201, China. E-mail: [email protected].
Abstract: Aiming at the problems of low accuracy and long time consumption of motion attitude information collection in the existing methods, a new automatic attitude capture method based on sensor information fusion is proposed. The line builds a motion attitude information acquisition architecture through five functional modules, including data acquisition and processing, laser sensor, data storage, voltage conversion and serial communication, to obtain motion attitude data. Then, the sensor image information collected by median filtering is processed, and the target image is obtained by fusing the binary image obtained by optical flow segmentation with the original image. Finally, the Mean-Shfit algorithm and particle filter are used to track the human moving target, and the self-similar matrix is used to automatically capture the motion posture. Experiments show that the method has high recall and accuracy for the collection results of motion attitude information, the accuracy rate of automatic motion attitude capture is as high as 95.5%, the maximum capture time is 79 ms, the time consumption is short, and the practical application effect is good.
