摘要
本文介绍了一种利用工业孔探仪对发动机内部损伤进行检测和识别的新方法。通过孔探探头采集发动机内部损伤图像,利用数字图像处理中的最大类间方差法分割出损伤区域,提取损伤图像的几何特征和纹理特征,并将提取的图像特征输入神经网络进行分层识别,最后由专家系统对损伤程度进行诊断。系统实现了孔探检测损伤的自动分类和损伤程度的自动诊断。通过现场测试,证明了该方法的有效性和实用性。
A new method is introducd to inspect and recognize engine faults with video borescope in this article. First, the engine＇s interior face fault was detected by video borescope, then the face fault was segmented by OTSU（ Maximization of interclass variance） algorithm from the borescope image. Second, three types of character parameters（surface texture, shape and size） were confirmed and used as the inputs of neural network to identify what kind of fault it is step by step. Finally, an expert system is adopted to judge the damage degree. By this method, danaeges can be classified and justified automaticly by means of images recognition, and the feasibility and validity of the method has been proved by the result of practice test.
出处
《节能技术》
CAS
2009年第1期69-73,共5页
Energy Conservation Technology
关键词
发动机损伤
孔探检测
数字图像处理
专家系统
engine fault
borescope inspection
digital image processing
expert system
