Fire Detection Based on Feature of Flame Color

Abstract

Color and its distribution are important features of a flame image. Fire detection based on sequences of images is more suitable for the need in big room or badly environment. A new color model used to detect flame in an image is found. The model is based on the features that were linear transformed from RGB of the image color. Flames in video sequences are detected by using the features of color and its distribution. Experiments show that the fire detection system is able to work well and get high detection rate with a low false positive rate.