摘要
为了准确分离目标区域,提出了一种基于贪婪Snake与多尺度图像增强相结合的新的轮廓分离模型.首先对输入的图像给定一个初始轮廓,然后应用多尺度分析进行图像增强,之后在不同的尺度中运用贪婪Snake模型进行轮廓的分离,最终分离出理想的目标轮廓.实验证明该方法提高了分割的精度.
Image segmentation is an important issue in the field of imaging and image processing.In order to separate target area accurately,a new method was proposed based on greedy Snake and muti-scale analysis to separate contour.An initial contour was given,image was enhanced by muti-scale analysis.The contour was then separated using the greedy snake model on different scales,to obtain an ideal contour.Experiments showed that this method improved segmentation accuracy.
出处
《北京师范大学学报（自然科学版）》
CAS
CSCD
北大核心
2011年第1期40-43,共4页
Journal of Beijing Normal University(Natural Science)
基金
国家自然科学基金资助项目(F020503/60803092)
北京市自然科学基金(4081002)
