Fault diagnosis of electrical distribution panel using thermal imaging
Loading...
Date
2019-06
Authors
Muhammad Yusri Bin Zaini
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Electrical fault is abnormal electric current where it is higher than the normal load. It can cause severe damage to the equipment and even loss of life. Thermal imaging camera has been introduced to do a preventive maintenance in the electrical power equipment. It can detect the temperature and thus hotspot region can be known. However, the manual analysis requires a lot of efforts, skills, time consuming and thus can lead to incorrect fault diagnosis. This is because the thermal images are usually low in resolution. They have a low contrast and contain background noise. Therefore, image processing techniques are used to analyze this thermal image. In this project, image filtration, image enhancement and image segmentation are used. Using MATLAB as a platform, the algorithm will process the thermal images. The hotspot region will be the region of interest and it will be segmented. The noises in the image are reduced by using Median filter, Wiener filter and Gaussian filter. The contrast and quality of the image will be enhanced by using histogram equalization technique. After that, it will calculate the threshold value of the images using modified Otsu’s method and compare the result with manual thresholding. Test results showed that hotspot region can be located easier after the quality of the image had been improved.