一种改进的基于小波变换的图像边缘检测算法  

A modified image edge detection algorithm based on wavelet transform

作　　者：范艳军[1] 吴晓红[1] 罗代升[1] 

机构地区：[1]四川大学电子信息学院,成都610064

出　　处：《四川大学学报（自然科学版）》2012年第6期1264-1268,共5页Journal of Sichuan University(Natural Science Edition)

基　　金：中央高校基本科研业务费专项资金(2009SCU11009);四川省科技支撑计划(10ZC0968)

摘　　要：本文研究了利用小波变换对图像进行边缘特征提取的算法.传统的小波边缘检测算法检测到的边缘均有边缘定位不准确、边缘不连续等缺点.为了解决这些问题,本文对小波边缘检测算法作了一些改进.主要改进了以下两个方面:在图像边缘检测方面,检测小波系数模的相角方向上小波系数的局部模极大值点;在阈值选取方面,不再应用全局的单一阈值,而是选择动态双阈值.实验表明,本文方法在图像边缘的连续性比较好,边缘定位上比较准确,而且能很好地保留图像的细节信息.This paper studies the wavelet transform algorithm which is used to detect image edge feature Classical wavelet algorithm of edge detection has these shortcomings of edge discontinuous and edge position inaccurate. In order to solve these problems, some corresponding modification for the wavelet edge detection algorithm are proposed to solve these problems in this paper. It is mainly improved the following two aspects： In the aspect of image edge detection, wavelet edge detection algorithm detectes local modulus maximum points of wavelet coefficients along the direction of the wavelet coefficients phase angle . In the aspect of threshold selection, instead of used the glabal single threshold, the dy- namic threshold is used for the selection of threshold. Experiments show that , the proposed method in this paper is precise in the image edge position, and the continuity of the edge is better, the detail infor- mation of the image could be reserved better.

关 键 词：小波变换 边缘检测 动态双阈值 

分 类 号：TN919.8[电子电信—通信与信息系统]