Result Analysis of Smoke Detection in Video for Early Warnings using Static and Dynamic Features

Abstract

paper presents result analysis of proposed method of smoke detection in video based on image processing to provide an early warning of anomalous events. The experimental results show that an adaptive background subtraction method, HSV color model and SVM classifier provides more accurate results in the segmentation and detection of smoke region in video. Also it shows that smoke features helps in decision making that is whether the segmented region in video is of smoke or not, hence this improves the reliability of system. The proposed method reduces the false detection and increases the smoke detection rate. This paper is organized into following sections. Section 1 contains introduction to smoke detection. Section 2 contains description of the proposed method. Section 3 contains result analysis of the proposed method. Section 4 contains the conclusion of the work.