一种图像轮廓数据的控制点检测算法

An Algorithm of Dominant Point Detection for Image Contour

针对如何提取图像轮廓控制点这一问题,提出了一种简单有效的方法。该方法分为去除坏点、标记轮廓数据点以及提取轮廓控制点3部分。首先计算轮廓数据点到拟合圆圆心的向量,然后根据向量间的关系找到坏点并去除,再对轮廓中的数据进行标记,最后对标记后的数据点进行数据缩减,并按照标记针对不同部分的数据提取轮廓控制点。试验验证提取出的轮廓控制点不仅能够表示轮廓的形状,也有效地消除了数据冗余。

A simple and efficient algorithm is presented for detecting dominant points of the image contour.The algorithm comprises identifying bad points,marking contour points and extracting contour dominant points.First the vector between the contour point and the center of the fitted circle are calculated.Then bad points are identified and removed according to the vectors′ relation,and contour points are marked based on some rules.Finally marked contour points are reduced and dominant points are extracted from different section′s data according to the mark.Experimental results show that the contour dominant points can not only represent the contour′s shape but also avoid redundant data.