Fire Boundary Detection Method Using a Unique Structure from Motion for Non-Rigid Bodies Algorithm (SFM-NRBA)

Abstract

Image processing has been widely used in various applications of object detections based on edges, motion and color detection. However edge, motion and color detection methods alone cannot accurately detect and identify events like the onset of fire in a scene. In order to detect and identify fires by using video we propose a technique using three steps search algorithms and structure from motion (SFM). Here in this paper we introduce a unique algorithm known as SFM for non-rigid bodies (SFM-NRB) algorithm. In this algorithm, we calculate a randomness figure of an object boundary to differentiate fire from none fire event. Novelty of this proposal is its ability to detect the onset of fires in an image by using a combination of simple three step search algorithm and randomness figure of object boundaries even under complex static and dynamic background conditions.