燕山大学学报2016，Vol.40Issue(5)：426-430,5.
基于局部SURF与Kalman滤波的多机器人识别与跟踪
Recognition and tracking of multiple robots based on local SURF and Kalman filter
摘要
Abstract
For improving SURF matching algorithm, an idea of local feature matching which can recognize and track multiple moving robots in one scene is proposed. In the static scenes, background subtraction method is employed to detect all the moving robots and extract the profile each of them. Through profile the hold region of each robot in the scene can be got. Extracting SURF feature points of each region and matching them with template image, the location information is achieved of the tracking robot. Finally, the Kalman filter is used to correct robot’ s location and track it. Experiments show that the proposed algorithm improves the efficiency of recognition and tracking without reducing the matching accuracy.关键词
SURF/特征匹配/背景差法/Kalman滤波Key words
SURF/feature matching/background subtraction method/Kalman filter分类
信息技术与安全科学引用本文复制引用
李江昊,韩立媛,彭丹,王绪全..基于局部SURF与Kalman滤波的多机器人识别与跟踪[J].燕山大学学报,2016,40(5):426-430,5.基金项目
国家自然科学基金资助项目（61403335）；河北省自然科学基金资助项目（F2013203186）；河北省高等学校青年基金资助项目 （）
