计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (20): 179-180.DOI: 10.3778/j.issn.1002-8331.2009.20.053

• 图形、图像、模式识别 • 上一篇    下一篇

结构件内部缺陷断层图像的模糊边缘检测

孔凡树,姚 芳,贾 超,邹 琪,王蓓蓓,艾 东   

  1. 燕山大学 信息科学与工程学院,河北 秦皇岛 066004
  • 收稿日期:2008-09-05 修回日期:2009-02-03 出版日期:2009-07-11 发布日期:2009-07-11
  • 通讯作者: 孔凡树

Fuzzy edge detection of inner defect based on cross-image

KONG Fan-shu,YAO Fang,JIA Chao,ZOU Qi,WANG Bei-bei,AI Dong   

  1. College of Information Science and Engineering,Yanshan University,Qinhuangdao,Hebei 066004,China
  • Received:2008-09-05 Revised:2009-02-03 Online:2009-07-11 Published:2009-07-11
  • Contact: KONG Fan-shu

摘要: 由于结构件内部缺陷形状复杂、随机性大及其断层图像噪声严重并具有一定的模糊性,在研究Pal模糊边缘检测算法的基础上,提出一种改进的模糊边缘检测算法。该算法将最佳阈值引进算法中,并以此改进了Pal算法中的隶属度函数,建立了新的模糊增强变换,最后增加了边缘连接的步骤。该算法具有较强的检测模糊边缘的能力,实验结果表明该算法是一种更实用、更高效的模糊边缘提取算法。

关键词: Otsu法, 隶属函数, 图像增强, 边缘连接

Abstract: Considering the main characters of inner defect in Industry Computer Tomography(ICT) image,such as complex shape,randomness,serious noise,and always fuzzy,this paper brings forward an improved fuzzy edge-detection method on the basis of studying the Pal algorithm.The algorithm introduces the best threshold into algorithm,and improves Pal membership function,then establishes a new fuzzy enhancement transformation,finally,an edge-connection step is added.The proposed algorithm is superior in detecting fuzzy edges.Experimental results show that the algorithm is a more practical,more efficient fuzzy edge extraction algorithm.

Key words: Otsu’method, membership function, image enhancement, edge-connection