摘要
在X射线检测焊缝缺陷过程中,至关重要的一步就是图象中缺陷的提取和分割,其处理效果直接影响到后续缺陷识别的正确性。本文先进行了焊缝背景去除,然后针对焊缝图象对比度差的特点以及其固有的模糊性,设计了基于直方图的模糊增强算法,以及基于直方图的模糊C-均值聚类分割算法。实验表明,本文提出的方法不仅能大大减少图象处理工作量,还能有效地提取出缺陷,为后续正确识别缺陷打下良好基础。
One important step of weld defect inspection by X-ray is how to extract and segment defects in X-ray inspection image. The processing result has effect on the correctness of next work for the defect recognition.Firstly,in this paper,the weld background is removed.Secondly,in the case of bad contrast ratio and fuzzyness of weld image,the vague strengthen algorithm and FCM algorithm based on histogram is designed.The experience shows that not only the amount of work is reduced largely,but also the defects is ex- tracted effectively by the method in this paper.lt is good base of defect recognition.
出处
《微计算机信息》
北大核心
2007年第18期300-301,286,共3页
Control & Automation
基金
2005年吉林省高等教育教学研究立项课题(吉林省教育厅:吉教高字[2005]23号)
关键词
焊缝图象
图象处理
模糊增强
缺陷提取
模糊聚类
Weld image
Image processing
Vague strengthen
Defect extraction
Vague assemble
