基于小波变换与灰色系统理论的图像边缘检测  

Image Edge Detection Based on Wavelet Transform and Gray System Theory

作　　者：郑子华[1] 陈家祯[1] 钟跃康[2] 潘日红[1] 

机构地区：[1]福建师范大学数学与计算机科学学院,福建福州350007 [2]福建师范大学信息管理系,福建福州350007

出　　处：《计算机与现代化》2010年第2期97-100,共4页Computer and Modernization

基　　金：福建省教育厅基金资助项目(JB09064);福建省教育厅基金资助项目(JB07051)

摘　　要：多尺度边缘综合一直是多尺度思想中较难解决的问题。本文将灰色系统理论与多尺度边缘检测思想相结合,提出基于灰色关联分析和梯度方向的图像边缘宽度计算策略,并根据边缘宽度自适应地调整样条小波的滤波尺度参数,从而实现对有噪图像进行自适应尺度边缘检测。仿真结果表明本文算法不仅能有效地抑制噪声,而且对细节边缘和模糊弱边缘均有较好的检测效果。Multi-scale edge synthesis is a difficult problem. Combining gray system theory with idea of multi-scale edge detection, an adaptive scale edge detection algorithm based on gray system theory is presented. It chooses scales of quadratic spline wavelet filter adaptively by the width of image edge which is calculated by a new method based on gray correlation analysis and Gradient Direction. The simulations prove that in binary image cases, the algorithm is good at removing noise and remaining both dim edges and fine edges.

关 键 词：边缘检测 灰色系统理论 灰色关联分析 小波变换 

分 类 号：TN911.73[电子电信—通信与信息系统]