ISBN: 978-981-11-0008-6 DOI: 10.18178/wcse.2016.06.137
Research and Application of Online Monitoring Fault Diagnosis for Infrared Image and Visible Image Registration
Abstract— To find the hidden trouble in substation accident early, it is necessary to build a infrared
temperature monitoring and warning system, which can collect infrared image of substation electrical
equipment automatically and circularly, through extracting the temperature information of the objective
device to distinguish fault. In order to solve the positional deviation of the image, which caused by the
mechanical transmission, leads to that it is difficult for systems to monitoring and positioning problem on
interested equipment, so it needs to use corrected method of image registration, firstly segmenting object and
defining configuration for image, and extracting the target which is detected, using suitable image
registration method to register the deviation between a sequence of image. So that the system can get the
temperature of the equipment parts in time and accurately, and ensure the reliable operation of the monitoring
system. The infrared image and the visual image registration need to be registered and the fault region in the
visible light image need to be marked for detecting the fault zone accurately. In the implementation, we use
the matching method for the registration of infrared image and visible light image which base on the trunk
line pair ofreference.
Index Terms— infrared temperature monitoring, positional deviation, image registration, trunk line pair.
Jianqiang Huang, Huan Wu, Xiaoying Wang, Zhengming Yi
State Key Laboratory of Plateau Ecology and Agriculture, Department of Computer Technology and Application, Qinghai University, CHINA
Cite: Jianqiang Huang, Huan Wu, Xiaoying Wang, Zhengming Yi, "Research and Application of Online Monitoring Fault Diagnosis for Infrared Image and Visible Image Registration," Proceedings of 2016 6th International Workshop on Computer Science and Engineering, pp. 765-772, Tokyo, 17-19 June, 2016.