Algorithm Research For Forest Fire Smoke Recognition

Forest fire is a natural disaster that threatens the ecological system seriously.For maintaining the ecological environment and reducing the economic loss,it is of great significance to realize the early warning of forest fire.Aiming to implement the intelligent early warning of forest fire,a method combined smoke and flame recognition was proposed for forest fire warning.For smoke recognition,the color features were used to determine the smoke suspected area.Besides,the texture features was utilized to extract the irregular,fuzzy,complex and correlative features from suspected area.For flame recognition,the color features were used to determine the flame suspected area.Besides,the texture and geometrical features were utilized to extract the features from suspected area.Texture features mainly include energy and contrast of gray level co-occurrence matrix(GLCM).Geometric features mainly include the computation of circularity and eccentricity.At last,the fire smoke was identified by support vector machine(SVM).The result demonstrated that the forest fire recognition method proposed herein provides an effective solution for the study of forest fire smoke recognition.