基于子形心集Hausdorf 距离的平面形状识别新方法
摘 要
本文在定义了平面形状边缘子形心概念的基础上，综合形状识别的Hausdorff距离法和形殊点法的思想,提出了一种形状分类新方法,并将它用于含刚性变化的封闭形状识别问题中.实验证明,此法在提高分类精度的同时使计算代价大大下降"较过去的形状识别方法来说更加快速有效.
关键词
Planer Shape Recognition based on Shape Subcenter Hausdorff Distance
() Abstract
Combined with shape-spccilic points and Hausdorff Distance,a new shape matching measurement based on the concept oI shape subcenters is given here.It is used for shape recognition under rigid transformations,brings about great improvements on precision.while decreases calculating cost very sharply. The test shows it's more effective and faster than any old method.
Keywords
|
中国图象图形学报 │ 京ICP备05080539号-4 │ 本系统由