摘要
脉冲涡流热成像检测技术是一种新型的无损检测方法,研究热成像表面缺陷的数据特征,将otsu分割算法与最大熵分割算法相结合,提出基于最大熵的otsu分割算法,既能较好的分割热图像中的缺陷又能有效的识别目标缺陷的边缘。实验结果表明,相比较其他传统分割算法,具有更好的缺陷提取效果。
Pulse eddy current thermal imaging detection technology is a new type of non-destructive detection method.It studies the data characteristics of thermal imaging surface defects,combines the otsu segmentation algorithm with the maximum entropy segmentation algorithm,and proposes the otsu segmentation algorithm based on the maximum entropy.Segmenting the defects in the thermal image can effectively identify the edges of the target defect.Experimental results show that compared with other traditional segmentation algorithms,it has better defect extraction effect.
作者
魏爱东
Wei Aidong(School of Automation Engineering,Shanghai University of Electrie Power,Shanghai,200090)
出处
《电子测试》
2020年第7期56-57,59,共3页
Electronic Test
关键词
分割算法
热图像
最大熵
无损检测
segmentation algorithm
thermal image
maximum entropy
NDT
