摘要
基于DR(航位推算,dead reckoning)算法的室外个人导航,应用三轴加速度传感器采集行人在运动过程中的数据,并对数据的特征进行分析,有效地识别出行人的水平行走和上、下楼梯、静止及跑等运动姿态。以步态检测为基础进行多种运动姿态下步长的计算,从而得到DR算法中的位移距离,以提高个人导航精度。研究将行人日常步态分为易分类(静止和跑)和不易分类(平地走、上楼及下楼),着重对不易分类的步态进行分类,最终通过小波分析及SVM(支持向量机)相结合的方式得到了精度较高的分类结果。
This paper focuses on the outdoor personal navigation based on dead reckoning algorithm. A three-axis acceleration sensor is used to collect pedestrian data during exercise, and feature data are analyzed to efficiently identify the level walk and on up or down the stairs, running and other sports static posture. The detection of gaits is used as the basis for calculation of more species, the displacement distances can be obtained by the DR algorithm, and used to improve the accuracy of personal navigation. In this paper, the pedestrian gait is divided into easy to classify(rest and run) and difficult to classify(go, go upstairs and downstairs). This paper focuses on the classification of gait that is difficult to classify. Finally, the results which are obtained by the combination of wavelet analysis and support vector machine are more accurate.
作者
马立
刘畅
MA Li;LIU Chang(Beijing Aerospace Petrochemical Technology And Equipment Engineering Corporation,Beijing 100076;University of Science and Technology Beijing,Beijing 100083,Beijing 100083)
出处
《微型电脑应用》
2019年第7期151-154,共4页
Microcomputer Applications
