摘要
文章提出了一种扩展的多尺度模糊边缘检测方法(IFED)。凭借小波多尺度理论去噪和准确定位的优势,算法将像素的邻域由8邻域扩展到了16邻域,有效地利用了多尺度滤波以及模糊竞争分类法则,较好地保留了图像的细节部分,提高了抗噪性。实验结果表明,与传统的边缘提取算法和模糊竞争算法相比,扩展的多尺度模糊边缘检测方法(IFED)能得到更满意的效果,尤其在抗噪性能上的结果是相当的理想。
This paper proposes improved multi-scale fuzzy edge deteetion（IFED）.Based on the advantages of less noise and accurate location in the multi-scale field,the novel algorithm extends eight-neighborhood to slxteen-neighborhood, applies multi-scale filter and competitive fuzzy rules to edge detection effectively,keeps the image details and improves the noiseproof feature.Compared with traditional edges and competitive fuzzy edge detection,IFED achieves better results,especially on the noiseproof feature facet.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第7期65-68,共4页
Computer Engineering and Applications
基金
重庆市自然科学基金资助项目(编号:CSTC2005BA2002)
关键词
邻域
多尺度
模糊竞争法则
边缘检测
pixel neighborhood,multi-scale,fuzzy competitive rules,edge detection
