Research of Fire Classification Algorithm

Wen Shen

Hua Qing College, Xi’an University of Architecture and Technology, Xi’an, 710045, China

Abstract: According to the flame image characteristics cannot be selected with scene Adaptive techquestions, this paper put forward a new algorithm which is based on Concept lattice and rough set, support vector machine (SVM), Through the descretization of fire flame characterstic datas, built the concept lattice gackground, calculate the difference matrix of form background, and use inportance index of characteristics to calculate the inportance of defferent characteristics. Finally put the minimalist feature classification sets into SVM to test. Experiments show that recognition rate of this method was obviously higher than that of single application of rough set for feature selection and feature selection yes artificial identification recognition rate, the efficiency is raised, false alarm is reduced.

Keywords: Image fire detection; Concept lattice; Rough set; Support vector machine