Enhanced image segmentation in thermal infrared image processing for faulty detection on broadcasting equipment / Mohd Rizman Sultan Mohd, Sukreen Hana Herman and Zaiton Sharif

Evolution of electronics equipment evolves
rapidly with the technologies invented. Broadcasting field
also transform from analogue devices system into file-based
system. As a continuously used equipment, the life-span is
an important issue. To prevent breakdown, this
broadcasting equipment must have a proper health
inspection which provides an early detection for further
action. This paper reports a faulty detection on
broadcasting equipment with thermal infrared imaging
system as a monitoring system. This study is carried out in
the radio broadcasting studio which monitor the equipment
through the operation time period. The data gathered from
the system will undergo enhanced image segmentation
method using k-means clustering with the implementation
of histogram equalization for further image processing
algorithm to detect the presence of hot spot. The final
output image is compared with the number of hot spots
detected for both images processed with implementation of
histogram equalization and the image without the
enhancement. Based from the result, it is shown that this
proposed method successfully improved the results of the
hot spot detection on thermal infrared images.