Monitoring and fault diagnosis of electrical Installation using thermal imaging
Loading...
Date
2018-06
Authors
Tiong, King Hock
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Electrical fault is any abnormal electric current bypass the normal load. Thermal
camera has been widely used for many years to monitor and inspect thermal defects of
electrical equipment by detecting the hot-spot. However, manual analysis of infrared images
for detecting defects and classifying the status of the equipment may consume a lot of time
and efforts, and may also lead to incorrect diagnosis. In order to tackle this problem, image
processing technique is introduced to analyze the possible electrical fault automatically from
infrared images. In this project, Otsu’s method, Wiener filter and histogram equalization are
implemented in analyzing the electrical fault. MATLAB is chosen as a platform to develop
the algorithm to process the infrared image. In the first stage, the infrared images will be
uploaded to the computer to analyze the high intensity parts of the image because these parts
represent higher temperature hot-spot. Next, the hot-spot regions of the image will be
cropped and the quality of these images will be enhanced by histogram equalization and
Weiner filter. Then, comparison between the image of normal current condition and image
of possible fault region will be carried out. After the comparison, it will alert the users if there
is any possible fault. In this project, the ordinary Otsu’s method is modified so that it is
suitable to be used to segment and analyse the electrical fault. If a fault is found on the system,
a message will be generated by the system to alert the user. In conclusion, the automatic
diagnosis technique can reduce a lot of manpower, time and prevent human error in electrical
fault diagnosis.