基于图搜索策略的边缘检测方法  

A method of edge detection based on the graph search strategy

作　　者：刘炀[1] 石文莹[1] 石鸽娅[1] 王静[1] 汤传玲[1] 

机构地区：[1]合肥工业大学,安徽合肥230009

出　　处：《微型机与应用》2012年第6期28-31,共4页Microcomputer & Its Applications

摘　　要：介绍了边缘检测的不同方法,并通过实验详细分析了不同算法的优劣。其中,边缘检测子依赖于灰度图像的边缘;内边界跟踪需要二值图像或是含有已经被标注出区域的图像;基于图搜索的边界跟踪通过获得相关启发性信息指导搜索以完成检测。重点讨论了采用不同启发式函数的边缘提取的优劣和效率。Firstly, the paper briefly introduces many kinds of the methods and species about edge detection. Secondly, the experiment results of these edge detection mentioned above are analyzed in details. Edge detector relies on the edge of gray image, and inner boundary tracking needs binary image, as well as image with marked regions. Graph search-based boundary tracing obtains correlating heuristic information by heuristic search, guiding search in order to complete detection. Thirdly, it puts emphasis on the results of edge abstract with different heuristic search function.

关 键 词：边缘检测子 内边界检测 图搜索策略 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]