本文已被:浏览 1655次 下载 2813次
Received:March 16, 2016
Received:March 16, 2016
中文摘要: 人的动作行为识别是计算机视觉领域中的热点研究课题之一,广泛应用于安防监控系统、医疗诊断监护、人机交互等方面.从运动对象检测、行为特征提取和人体行为识别方法方面阐述了人体行为识别技术的现状和相关难点.与已有相关综述性文章相比,结合了新的研究成果,介绍了基于轨迹能量扩散图的组行为识别方法以及基于稠密光流轨迹和稀疏编码算法的行为识别算法.且指出了未来行为识别的发展趋势.
Abstract:Human motion recognition is one of the hot topics in the field of computer vision,which is used widely in security monitoring system,medical diagnosis and monitoring,human-computer interaction,etc.The current state and related difficulties of human motion recognition technology is presented within the motion object detection,feature extraction and human motion recognition method.Compared with the existing relevant review articles,the new research results are combined,such as group behavior recognition method based on trajectory energy diffusion map and behavior recognition algorithm based on dense flow trajectory and sparse coding algorithm.Furthermore,some research direction of motion identification on future work is given.
文章编号:20171022 中图分类号: 文献标志码:
基金项目:
Author Name | Affiliation | |
ZHANG Xingjian | School of Electronics and Information Engineering, Shanghai University of Electric Power, Shanghai 200090, China | txhtxh1991@163.com |
Author Name | Affiliation | |
ZHANG Xingjian | School of Electronics and Information Engineering, Shanghai University of Electric Power, Shanghai 200090, China | txhtxh1991@163.com |
Reference text: