东南大学学报（自然科学版）2011，Vol.41Issue(2)：266-269,4.DOI:10.3969/j.issn.1001-0505.2011.02.010
多媒体传感器网络中被动目标定位算法
Passive target localization algorithm for multimedia sensor networks
摘要
Abstract
A multi visual sensor node collaborative localization algorithm is proposed for the passive target positioning problem. Different from traditional localization methods based on vision, this method doesn't need training or feature recognition. Based on machine vision theory, every node projects the target to be a straight line on the ground plane, and then the line is mapped to be a discrete point in Hough spaces using Hough transform. Finally these points are fitted to be a sine curve, and the fitting parameters are the target position estimation. In addition, the accuracy of multi visual sensor node collaborative localization algorithm is further improved by introducing confidence parameter based on the height and resolution of visual sensor node. Finally, time complexity is analyzed, and experiments confirm that this algorithm can obtain a localization accuracy better than 5 cm.关键词
多媒体传感器网络/协作定位/Hough变换/正弦曲线拟合/最小二乘Key words
multimedia sensor networks/ collaboratively localization/ Hough transform/ sine curvefitting /least squares分类
信息技术与安全科学引用本文复制引用
张波,罗海勇,胡广大,刘冀伟,赵方..多媒体传感器网络中被动目标定位算法[J].东南大学学报（自然科学版）,2011,41(2):266-269,4.基金项目
国家自然科学基金资助项目(60873244,60973110,60772070)、国家高技术研究发展计划(863计划)资助项目(2009AA011902)、北京市自然科学基金资助项目(4102059)、工业和信息化部重人专项资金资助项目(2010ZX03006-002-03). （60873244,60973110,60772070）
