快速图像处理中阈值选取方法的比较研究  

A Comparative Study of Threshold Selection Methods in Fast Image Processing

作　　者：李卓[1] 郭立红[1] 

机构地区：[1]吉林长春中国科学院长春光学精密机械与物理研究所

出　　处：《微计算机信息》2006年第03S期224-225,163,共3页Control & Automation

基　　金：国家863计划(863-703主题基金)

摘　　要：图像分割是图像处理中的一个重要问题。阈值法计算简单,具有较高的运行效率,这使得其在实时图像处理系统中得以广泛的应用。本文选取在实时性强的系统中使用较好的几种阈值化方法进行分析,阐述了各种方法的基本思想,以及各自的优缺点和适用范围,并对相应的改进算法进行了综述,对进行图像分割研究的科研人员具有一定的参考价值和启发作用。Image segmentation is important in digital image processing. Thresholding methods are widely used especially in real-time image processing systems because of their simplicity and efficiency. This paper analyses a few methods for threshold selection in real-time systems. And the basic ideas of various methods as well as their advantages and defects are critically reviewed. The improved algorithms are also summarized. This paper is valuable and useful for the researchers in the image segmentation field.

关 键 词：快速图像处理 阈值化 直方圈 图像分割 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]