基于评价知识的图象分割算法优选系统  

Evaluation Knowledge Based Optimal Selection System for Image Segmentation Algorithms

作　　者：章毓晋[1] 罗惠韬 

机构地区：[1]清华大学电子工程系

出　　处：《高技术通讯》1998年第4期21-24,共4页Chinese High Technology Letters

基　　金：国家教委博士点基金;863计划资助项目

摘　　要：报道了基于评价知识对分割算法进行优选的研究进展。提出了通用分割评价框架和客观定量的分割质量评价准则，据此利用对算法进行评价和对分割效果进行比较的方法实现了第一个图象分割算法优选系统。讨论了基于评价知识选取最优分割算法的策略和系统设计的原理，给出了用实际图象进行算法优选和分割实验得到的结果，验证了系统的有效性。The research progress in the optimal selection of segmentation algorithms using evaluation kowledge is reported in this paper. First, the proposed general framework for segmnetation evaluation and the objective and quantitative criteria for segmentation quality judgment are introduced. Then, the first optimal selection system based on the evaluation of algorithms and the comparison of segmentation results is developed. The strategy of selecting the best algorithm based on evaluation knowledge and the principal of system design are discussed. In addition, the results of algorithm selection and the segmentation experiments with real images are provided to show the effectiveness of the system.

关 键 词：图象分割 分割评估 评价知识 算法优选 图象处理 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TN919.8[自动化与计算机技术—计算机科学与技术]