基于特征点匹配的全局运动估计  

Global Motion Estimation Based on Feature Points Matching

作　　者：武艳美[1,2,3,4] 肖阳辉[2,3,4] 

机构地区：[1]中国科学院研究生院,北京100049 [2]中国科学院沈阳自动化研究所,沈阳110016 [3]中国科学院光电信息处理重点实验室,沈阳110016 [4]辽宁省图像理解与视觉计算重点实验室,沈阳110016

出　　处：《计算机工程》2011年第22期148-150,共3页Computer Engineering

摘　　要：传统块匹配法只考虑平移运动,对大角度旋转无法进行有效估计。为此,根据圆形的旋转不变性,提出一种基于圆形块的匹配方法,提取圆投影矢量作为匹配特征,采用特征匹配的策略,得到当前块的运动矢量。为提高匹配效率,利用Harris算子检测特征点,获取特征点的运动矢量。通过局部运动矢量和刚体变换模型得到线性参数方程,并采用RANSAC算法进行求解,以获得全局运动参数。实验结果证明,该方法对较大的平移及旋转角度具有较强的适用性。The traditional block matching method only considers the translation motion.It can not estimate the rational motion correctly.Considering the circle is robust to rotation,a circular block matching method is proposed.It adopts the circle projection as the feature vector,and obtains the motion vector of the current block.In order to improve the efficiency,it adopts the feature points using Harris arithmetic operator to get the motion vectors.A linear equation is constructed from the local motion vector and the rigid model,and it determines the global motion parameters by the popular RANSAC algorithm.Experimental results show the method is robust to determine the global motion parameters even if they are very big or small.

关 键 词：圆投影 圆形块匹配 角点 RANSAC算法 全局运动估计 

分 类 号：N945[自然科学总论—系统科学]