TCAS板卡红外诊断系统中红外图像分割研究  

The Infrared Image Segmentation Research of the Infrared Fault Diagnosis System of the TCAS Circuit Board Process

作　　者：王力[1] 高博[1] 程宝英[1] 

机构地区：[1]中国民航大学航空自动化学院,天津300300

出　　处：《控制工程》2013年第3期475-477,488,共4页Control Engineering of China

基　　金：国家自然科学基金(81000639);中央高校基本科研业务项目(ZXB2011A003);机载电子系统深度维修测试通用平台研究:中国民航大学科研基金(2010kyE07);基于红外图像处理的TCAS系统深度维修测试系统研制

摘　　要：传统的Otsu算法对于非双峰直方图图像分割有很大局限性,采用改进的基于迭代自组织(ISODATA)聚类法的Otsu算法对TCAS板卡的红外热像进行分割处理,首先利用Otsu算法计算分割阈值,然后利用ISODATA聚类法求出Otsu阈值算法的目标函数的标准差,将板上的电子元器件清晰的分割出来,为后续的故障诊断提供有效的保证。仿真结果表明,相对于传统的Otsu算法,这种方法针对TCAS板卡红外热像的图像分割更为有效。The traditional Otsu algorithm for non Shuangfeng histogram of image segmentation has great limitations, Through adopting the improved Otsu algorithm based on ISODATA（ Iterative Self-organizing Data Analysis Technique）, the segmentation processing of infrared image of the TCAS circuit board is achieved. First using Otsu to compute segmentation threshold , And then using ISODATA clustering method tocalculator the objective function of the standard deviation of the Otsu threshold algorithm. The electronic components of the boards are divided up clear, which is a guarantee for the subsequent fault diagnosis. Simulation results show that the method of the image segmentation of TCAS circuit board is more effective.

关 键 词：红外热成像技术 图像分割 迭代自组织聚类法 OTSU算法 TCAS板卡 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]