基于oghms及isef的红外图像边缘检测方法的研究-通信与信息系统专业论文.docxVIP

中山大学硕士学位论文 Title：Study on edge detection method of infrared image based on oGHMs and ISEF Major：Communications＆Information Systems Name：Yue Xiuyu Supervisor：Professor SHEN Wei ABSTRACT Fault diagnosis and the alarm are the core section ofthe infrared image monitoring system in the unattended substations It is important for power system to monitor the online temperature of electrical equipments by the infrared diagnosis technology Among the research，most of the functional softwares couldn’t inspect the faults automatically at present，and there are still lots ofproblems needing to be solved In this papeL it is the main problem how tO separate the faults region in the infrared image effectively The segment result is the precondition for the research of the fault diagnosis．As is well known，edge is the primary characteristic of the image segmentation，which includes the useful information of the image，Thus edge detection is kernel of segmenting the faults region in the infrared image． The principle of the edge detection presented by J Shen is analyzed in detail and three classic algorithms arc discussed．Based on this study,we propose an approach that the orthogonal Gaussian—Hennite moments integrate with the infinite symmetric exponential filter Applying the approach，the fault region is segmented effectively Then according to the Technical guide for infrared diagnosis of alive equipment，the hidden faults would be inspected by the temperature distribution．Experimental results demonstrate that the performance of our approach is better-performing on infrared image．Furthermore，the project is feasible that hidden faults of electrical equipments in the unattended substations could be inspected by the edge detection technology． Keywords：infrared image，edge detection，infinite symmetric exponential filter orthogonal Gaussian—Hermite moments  基于0GHMs及ISEF的红外图像边缘检测方法的研究 第1章绪论 1．1课题研究的背景及意义 1．1．1研究红外诊断电气设备故障的意义 随着社会经济的进一步发展，电力工业也在向着大容量高电压供电方向迅速 发展，随之而来，就对供电的可靠性提出越来越高的要求。在电力系统的生产过