首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于图像识别的发动机内窥智能检测系统研究
引用本文:李岩,孟娇茹.基于图像识别的发动机内窥智能检测系统研究[J].中国民航学院学报,2005,23(4):15-19.
作者姓名:李岩  孟娇茹
作者单位:[1]中国南方航空北方公司飞机维修基地,辽宁沈阳110043 [2]黑龙江科技学院自动化系,黑龙江哈尔滨150027
基金项目:中国民用航空总局科技基金项目(03126).
摘    要:介绍了一种基于图像识别技术的航空发动机内窥检测的新方法,采用最大类间方差法计算二值化分割的最佳阈值,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了诊断。通过现场测试,证明了该方法的有效性和实用性.

关 键 词:图像识别  内窥检测  特征提取  BP神经网络  专家系统
文章编号:1001-5000(2005)04-0015-05
收稿时间:2005-06-06
修稿时间:2005-06-20

Study of Intelligent Video Probe Inspection System for Aero-Engines Based on Image Recognition Technology
Li Yan;Meng JiaoRu.Study of Intelligent Video Probe Inspection System for Aero-Engines Based on Image Recognition Technology[J].Journal of Civil Aviation University of China,2005,23(4):15-19.
Authors:Li Yan;Meng JiaoRu
Abstract:Introducing a new video-probe-inspection method based on image recognition inside engines of the mechanical damages. With the superior threshold of image binarization by OTSU(maximization of interclass variance) algorithm and a method to extract damage characteristics,the extracted image characteristics is inputted into neural network to identify step by step what kind of damage it is,and an expert system is involved to do a diagnosis and assess the damage degree. An automatic classification of inside-vision damages and automatic diagnosis and verification of damages' extent is therefore brought about. Its effectiveness and reliability are testified through an on-the-spot experiment.
Keywords:image recognition  video probe inspection  characteristic extraction  BP neural network  expert system
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号