基于小波模极大值和形态学的图像边缘检测算法  

EDGE DETECTION APPROACH FOR IMAGES BASED ON WAVELET MODULUS MAXIMUM AND MORPHOLOGY

作　　者：胡艳[1] 张瑞林[1] 

机构地区：[1]浙江理工大学软件工程研究所,浙江杭州310018

出　　处：《计算机应用与软件》2010年第9期75-76,142,共3页Computer Applications and Software

基　　金：浙江省自然科学基金资助项目(ZJ602014)

摘　　要：提出一种基于小波变换和形态学的图像边缘检测方法。通过对源图像进行小波分解,用小波模极大值法和基于数学形态学的算法分别提取高低频子图像的边缘,最后采用合理的融合规则将两个边缘图像进行融合。实验结果表明,该算法能有效地抑制噪声,且边缘清晰、准确,效果优于经典的边缘检测算法。A method combining wavelet transform and morphology for solving image edge detection problem was presented. The method de- composes original image with wavelet transform, then uses modulus maximum of wavelet transform and mathematical morphology based algo- rithm respectively to extract the edges in high-frequency and low-frequency sub-images. Finally, the method adopts reasonable fusion rule to fuse two edge images. Experimental results show that the algorithm can effectively suppresses noise, the edges are clear and accurate, the effect outperforms the classical edge detection algorithm, so it is feasible and effective.

关 键 词：边缘检测 小波变换 数学形态学 图像融合 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]