###
计算机系统应用:2018,27(11):224-230
←前一篇   |   后一篇→
本文二维码信息
码上扫一扫!
基于拉班空间的跌倒方位检测方法
徒鹏佳1, 李军怀1,2, 王怀军1,2, 姬文超1,2, 王侃1,2
(1.西安理工大学 计算机科学与工程学院, 西安 710048;2.陕西省网络计算与安全重点实验室, 西安 710048)
Falling Position Detection Method Based on Laban Space
TU Peng-Jia1, LI Jun-Huai1,2, WANG Huai-Jun1,2, JI Wen-Chao1,2, WANG Kan1,2
(1.School of Computer Science and Engineering, Xi'an University of Technology, Xi'an 710048, China;2.Shaanxi Key Laboratory for Network Computing and Security Technology, Xi'an 710048, China)
摘要
图/表
参考文献
相似文献
本文已被:浏览 40次   下载 68
投稿时间:2018-04-12    修订日期:2018-05-08
中文摘要: 研究跌倒检测方法对于保障老年人健康生活具有重要意义.首先通过惯性传感器对老年人日常行为及跌倒动作进行捕捉,使用欧拉角法表示动作数据,引入滑动平均滤波算法对数据降噪融合;在此基础上,基于拉班舞谱空间表示方法,建立跌倒及方位检测模型,基于该模型进行异常姿态检测及八种精确方位判定,最终提出一种基于拉班空间的跌倒方位检测方法.实验结果表明,该方法的检测准确率可达100%,同时能够对老年人跌倒后受伤部位进行预诊断及报警.
Abstract:In moving object detection process, it needs to automatically judge whether it has detected the moving object, although there is no moving object in the current scene, detection result wrongly judges that it has detected the moving object. In order to find the source of the error, optical flow disturbance effect is found through experiment. The optical flow disturbance effect detection algorithm is designed, and the effect of optical flow perturbation is clearly detected. Next, through the binarization method of image it eliminates optical flow disturbance effect. The ideal results of the moving object detection are obtained. This research proves that the optical flow perturbation effect exists in the space, which can cause interference to the detection of moving object. It also can eliminate the effect of optical flow disturbance and improve the accuracy and reliability of moving object detection and judgment.
文章编号:     中图分类号:    文献标志码:
基金项目:国家自然基金(61771387);国家重点研发计划(2017YFB1402103);赛尔网络创新项目(NGⅡ20150707,NGⅡ20160704)
引用文本:
徒鹏佳,李军怀,王怀军,姬文超,王侃.基于拉班空间的跌倒方位检测方法.计算机系统应用,2018,27(11):224-230
TU Peng-Jia,LI Jun-Huai,WANG Huai-Jun,JI Wen-Chao,WANG Kan.Falling Position Detection Method Based on Laban Space.COMPUTER SYSTEMS APPLICATIONS,2018,27(11):224-230

用微信扫一扫

用微信扫一扫