基于边缘链码滤波的角点角度快速估计算法  

A New Method for Estimation of the Corner Angle

作　　者：成兰[1] 田原[2] 梁德群[3] 王红光 鲍亚萍 

机构地区：[1]河南安阳大学电气工程系,河南安阳455000 [2]中国科学院自动化研究所模式识别国家重点实验室,北京100080 [3]西安交通大学图像处理与识别研究所,陕西西安710049

出　　处：《华北电力大学学报（自然科学版）》2000年第4期59-63,共5页Journal of North China Electric Power University：Natural Science Edition

摘　　要：角点的角度是衡量角点特征的一个极为重要的指标，常作为阈值决定角点的取舍。本文提出一种新的角点角度估计算法，算法中通过对角点边缘链码采用不同方差的高斯函数进行卷积，得到角点位置的偏移，并给出由此偏移量计算角点角度值的公式。此算法具有较小的计算量。实验证明，此算法在一定场合下能有效地完成对角点角度的估计。文中最后给出了估计的误差分析。The comer feature has significant role in the computer vision. The angle is one of the most important features of the corner, which is often used as a threshold to decide whether this corner can be utilized. In this paper, a new algorithm for estimating of the comer angle is proposed. Firstly the algorithm convolutes the code chain of edge with different mean square in the neighbor of the corner. Therefore the different displace of corner can be obtained and then the angle of corner will be estimated by using this displace. The algorithm is less complex in computation and can estimate corner angle accurately in the proper field. The analyses of computing errors are also shown in the end of the paper.

关 键 词：计算机视觉 角点角度估计 边缘链码 滤波 算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]