计算机应用与软件2018，Vol.35Issue(2)：235-241,7.DOI:10.3969/j.issn.1000-386x.2018.02.043
基于下肢轮廓角度距离和步幅变化的步态识别算法研究
RESEARCH OF GAIT RECOGNITION ALGORITHM BASED ON LOWER-LIMB CONTOUR POINT DISTANCE AND STRIDE ALTERATION
王希娟1
作者信息
- 1. 陕西国际商贸学院信息与工程学院 陕西西安 712046;西安科技大学计算机科学与技术学院 陕西西安 710054
- 折叠
摘要
Abstract
In the video surveillance system for real-time gait recognition,we not only need to ensure high recognition rate,and need to minimize the recognition time.Therefore, this paper proposed a gait recognition algorithm based on lower-limb contour point distance and stride alteration.Firstly,the extracted contour image was cropped according to its minimum circumscribed rectangle,and the image was scaled according to the maximum height of all the images while keeping the aspect ratio unchanged.Then,the centroid height of all the image frames in a period was extracted,and the rate of change of the angle of the lower extremity contour between the two adjacent frames based on the pelvic joint point was taken as the gait feature.Finally, the similarity measure and the K-nearest neighbor algorithm were respectively completed by the dynamic time warping(DTW)algorithm and the K-Category identification.The experimental results showed that the proposed algorithm balanced the recognition rate with the time cost,and achieved high recognition rate of user identity in a short time,and had good robustness to clothing and backpack status.关键词
步态识别/下肢轮廓/角度距离变化/步幅变化Key words
Gait recognition/Lower-limb contour/Point distance alteration/Stride alteration分类
信息技术与安全科学引用本文复制引用
王希娟..基于下肢轮廓角度距离和步幅变化的步态识别算法研究[J].计算机应用与软件,2018,35(2):235-241,7.
