基于二维小波的粒型图像边缘检测  

Edge Detection Based on Two-dimension Wavelet for a Grain-image

作　　者：李葆青[1] 文山[1] 

机构地区：[1]六盘水师范高等专科学校,贵州六盘水553004

出　　处：《六盘水师范高等专科学校学报》2004年第6期5-8,共4页Journal of Liupanshui Teachers College

摘　　要：针对粒型图像感兴趣的部分,即粒形的径粒大小、面积、周长、分布特征等参量,要求进行精确的边缘检测,提出一种新的判断和检测出粒型图像边缘的闭合单线条方法,该方法采用二维小波对图像进行预处理,哈夫(Hough)变换帮助作闭合性判断,进而分析得到闭合的单线条边缘。实践证明,通过该方法边缘检测后得到的图像有着比较细的边界,而且边界是闭合的,图像清晰明了,便于更进一步的分析。Because of being interested in a certain patr of grain-imge,such as their dimension,area,perimeter and distributive characteristic,it is required to obtain an accutately detected edge.This paper offers a new method that can help to judge and check out a closed single-strip grain-image edge.In order to make the algorithm sententious and effective,this method adopts two-dimension wavelet to carry out the pretreatment in the image.In the meantime,Hough Transform helps to do closure judges so that we can analyze and get a closed single-strip edge.It has been proved by practice that this method produces a thinner boundary,and the boundary is a close while the inage appears to be sharp;therefore,it is easy for us to do further analysis.

关 键 词：小波变换 粒型图像 边界闭合 边缘检测 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]