基于阈值法的图像分割技术  

Image Segmentation Technology Based on Thresholding

作　　者：阴国富[1] 

机构地区：[1]西安电子科技大学

出　　处：《现代电子技术》2007年第23期107-108,共2页Modern Electronics Technique

基　　金：渭南师范学院科研基金资助项目(07YKS025)

摘　　要：图像分割是图像处理方向的一个重要课题,阈值法因其实现简单、计算量小、具有较高的运算效率,性能较稳定,而成为图像中最基本和最广泛的分割技术。阈值法分为全局阈值法和局部阈值法两种,其中全局阈值法又可分为基于点的阈值法和基于区域的阈值法。这里主要研究基于点的全局阈值法,即最大类间方差法,并基于Matlab软件环境进行算法的仿真。该系统能够对图像中某些感兴趣的部分进行提取,较好地实现了图像分割。The thresholding is a traditional image segmentation method, because of the simple realization, small computation load,having high operation efficiency as well as stable performance,the method becomes the most basic and widespread technology in image segmentation. The thresholding divides into two kinds as the global thresholding,and the global tresholding divides into methods based on point and region. This article mainly studies two kinds of global thresholding based on point, respectively is,OTSU thresholding and maximum entropy method, realizes two algorithms by software, carries on simple morphology process to the processing result and analysis result of image segmentation. This system can help people abstract something in which they interested.

关 键 词：图像分割 最大类间方差法 阈值法 动态阈值 

分 类 号：TN919.8[电子电信—通信与信息系统]