Vision Based Wildfire Detection Using Bayesian Decision Fusion Framework

Abstract: Computational vision-based fire and flame detection has drawn significant attention in the past decade with camera surveillance systems becoming ubiquitous. Several signal and image processing methods are developed for the detection of fire, flames and smoke in large and open spaces with a range of up to 30 meters to the surveillance camera in visible-range video. This paper proposes a new approach to vision-based wildfire smoke detection by using a compound algorithm and a decision fusion framework with Bayesian classifier as classification tool. The compound algorithm is a combination of several sub-algorithms, the fusion network is to fuse the outputs obtained by each of these sub-algorithms and finally a Bayesian classifier is used for distinguishing fire regions from non-fire regions. This technique is to improve the accuracy of wildfire smoke detection in videos and to reduce the false alarm rate to a great extent.

Keywords: Computer vision, Bayesian decision fusion, feature extraction, fire detection, generic colour model, image processing.