苑晶,李阳,董星亮,黄亚楼.基于运动模式在线分类的移动机器人目标跟踪[J].仪器仪表学报,2017,38(3):568-577
基于运动模式在线分类的移动机器人目标跟踪
Target tracking with a mobile robot based on on line classification for motion patterns
  
DOI:
中文关键词:  移动机器人  人体跟踪  单目视觉  运动模式分类
英文关键词:mobile robot  human tracking  monocular vision  classification of motion patterns
基金项目:国家自然科学基金项目(61175085,61175083,61573196)、天津市自然科学基金项目(15JCYBJC18800)、中国民航信息技术科研基地开放课题基金项目(CAAC-ITRB-201503)项目资助
作者单位
苑晶 1.南开大学计算机与控制工程学院天津300071;2.天津市智能机器人技术重点实验室天津300071;3.中国民航大学 中国民航信息技术科研基地天津300300 
李阳 南开大学计算机与控制工程学院天津300071 
董星亮 南开大学计算机与控制工程学院天津300071 
黄亚楼 1.南开大学计算机与控制工程学院天津300071;2.天津市智能机器人技术重点实验室天津300071 
AuthorInstitution
Yuan Jing 1. College of Computer and Control Engineering, Nankai University, Tianjin 300071, China;2. Tianjin Key Laboratory of Intelligent Robotics, Tianjin 300071, China; 3. Information Technology Research Base of Civil Aviation Administration of China, Civil Aviation University of China, Tianjin 300300, China 
Li Yang College of Computer and Control Engineering, Nankai University, Tianjin 300071, China 
Dong Xingliang College of Computer and Control Engineering, Nankai University, Tianjin 300071, China 
Huang Yalou 1. College of Computer and Control Engineering, Nankai University, Tianjin 300071, China;2. Tianjin Key Laboratory of Intelligent Robotics, Tianjin 300071, China 
摘要点击次数: 695
全文下载次数: 0
中文摘要:
      区别于以往采用固定运动模式的目标跟踪研究,提出一种基于单目视觉传感器的人体运动模式在线识别算法,及基于此算法的人体目标跟踪方法。首先,利用视觉信息检测运动目标,并提取其视觉特征;然后通过单目视觉深度提取算法,获取目标的运动特征;接着将连续几帧的特征变化矢量送入随机森林(RF)进行学习,实现对人体运动模式的在线分类;最后根据分类结果在线选取不同的目标运动模型,并利用近似最优的粒子滤波器实现对目标运动状态的准确估计。实验结果证明了本文提出算法的有效性。
英文摘要:
      Different from the existing researches on target tracking using a fixed motion pattern, an algorithm is proposed for on line recognition of human motion patterns with a monocular camera, to develop a human tracking method. First, moving target is detected using visual information and the visual features are extracted. Then, the motion features of target are acquired by depth extraction algorithm of monocular vision. The differences of a couple of successive frames' features are fed into random forest (RF) classifier to recognize human motion patterns online. Finally, different target motion models are chosen online based on classification results and approximate optimal particle filter is used to realize accurate estimation of human’s states. Extensive experimental results demonstrate the effectiveness of the proposed algorithm.
查看全文  查看/发表评论  下载PDF阅读器