摘要
以单目视觉为模型给出基于视觉的移动机器人目标定位系统算法,利用尺度不变特征变换匹配方法(SIFT)求得基本矩阵与本质矩阵,分解本质矩阵后进一步得到相机的运动参数,最后结合立体视觉方法以及运动恢复结构方法获得目标的深度信息,实现目标定位。给出了三维信息获取过程与计算机仿真结果。实验结果表明,算法能满足移动机器人对目标实现定位的精度要求。
A vision-based mobile robot localization method was proposed. The fundamental matrix and essential matrix was obtained by Scale Invariant Feature Transform (SIFT) matching algorithm. After the decomposition of essential matrix, the external motion parameters of single-eye camera were got. Finally, the structure was used from motion method to obtain three-dimensional information and carry out object localization. The three-dimensional information acquisition process and the computer simulation result were given. The results show that the method can meet the demand of robot localization manipulation precision.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2012年第9期1823-1825,1830,共4页
Journal of System Simulation
基金
国家自然科学基金(60872076)
高科技研究(863)(2006AA09Z203)
湖南省科技计划(2010GK3012)
湖南省青年骨干教师资助项目
关键词
单目视觉
目标定位
基本矩阵
尺度不变特征变换
single-eye vision
object localization
fundamental matrix
Scale Invariant Feature Transform