摘要
基于建立两个平面形状特征点之间的对应关系的问题上,提出一种新的解决方法。这种方法是把特征点从源形状和目标形状中抽取出来,根据此方法来衡量两个特征点之间相似度,采用有效的动态编程的方法找出特征点之间的最优化的对应关系来解决问题。该算法简洁、快速,并有效地保留原始形体上的特征属性。
Based on the problem constructing the corresponding relations between the feature points of two flat shape,puts forward a new method.This method puts the feature point from source shape and target shape extracted,and based on this method,measures the similarity between two feature points,adopts effective dynamic programming method of feature points and finds out the corresponding relationship between optimization to solve the problems.This algorithm is simple,fast and effectively retains the primitive physique characteristic attributes.
出处
《现代计算机》
2011年第9期34-38,49,共6页
Modern Computer
关键词
特征点
相似度
代价函数
插值
Feature Point
Similarity
Cost Function
Interpolation