Optimized Fire Detection on Surveillance System using Image Processing and Convolutional Neural Networks

Abstract

Nowadays fire outbursts are the most common issues which has immense effect on nature as well as human lifestyle. Due to this the urge for efficient fire detection applications has been increased. Hence in this proposed work a novel technique for early detection of fire on surveillance systems using various image processing techniques are discussed. Since the traditional image processing algorithms come up with some limitations like high false recognition of fire, the proposed work compares the performance of conventional image processing algorithms with convolutional neural network techniques. Convolutional neural network (CNN) is a deep learning algorithm which has prodigious performance in the field of object identification and classification.