摘要
脉冲涡流热成像是无损检测的重要方法之一。及时发现构件缺陷,能够防止事故发生。最大类间方差分割方法是提取脉冲涡流热成像缺陷的有效方法。同态滤波能够增强图像,提高缺陷分割效果。实验证明基于同态滤波和改进最大类间方差分割方法的缺陷提取效果较好。
Pulse eddy current thermal imaging is one of the important methods of nondestructive testing.Defect extraction of pulsed eddy current thermal imaging images can detect component defects in time and prevent further expansion of the accident.The segmentation method based on maximum inter-class variance is an effective method for extracting defects of pulsed eddy current thermal imaging.Homomorphic filtering can enhance images and improve defect segmentation.Experiments prove that the defect extraction based on homomorphic filtering and improved maximum variance between classes is better.
作者
车倩倩
Che Qianqian(College of Airtomotion Engineering Shanghai University of Electric Power,Shanghai,200090)
出处
《电子测试》
2020年第7期58-59,共2页
Electronic Test
关键词
同态滤波
最大类间方差分割
缺陷提取
Homomorphic filtering
maximum inter-class variance segmentation
defect extraction
