一种基于主动轮廓模型的运动目标跟踪方法

分析了几种经典主动轮廓模型（即Snake）的优缺点．针对传统Snake在应用于目标跟踪时的一些缺陷，设计了自动初始化轮廓的算法．采用四邻域交替搜索法取代八邻域搜索方法以减少计算开销，并给出了一种较为实用的基于Snake的运动目标跟踪方法．通过实验证明了本方法的可行性．

A Moving Object Tracking Method Based on Active Contour Models

This paper discusses the advantages and disadvantages of several representative active contour models(i.e.,Snakes).When traditional Snakes are used in moving object tracking applications,they often encounter some defects,and an automatic initialization algorithm is designed to overcome these problems.The eight-neighbor searching algorithm is replaced by the four-neighbor-alternating searching method to reduce the computing cost,and a practical moving object tracking method based on Snake is proposed.Experiments are made to demonstrate the feasibility of the presented method.