Detecting ROIs in the thermal image of electrical installations

This paper presents a method of detecting and segmenting regions of interest (ROIs) of the thermal image of electrical installations. These regions are very important in diagnosing the thermal condition of electrical equipment. Due to the nature of thermal imaging, segmentation with the conventional approach will make inaccurate ROI detection, especially when qualitative approach is considered in evaluating the equipment´s condition. Therefore, we take the advantage of extracting local features to identify, locate, and match multiple repeated objects and grouping a look-like similar objects in the images. Experimental results show that the proposed method achieves better performance for detecting the target ROIs with various irregular intensity variations, dim target equipments and cluttered background. The performance of the proposed method is qualitatively and quantitatively evaluated.