关键词: 火焰探测, 颜色模型, 小波变换, 支持向量机

Abstract: In order to improve fire detection adaptive ability, this paper proposes a video fire detection algorithm based on wavelet transform and support vector machine. Candidatefire regions are detected by the color of the flame probability model, and the features are the area changing, the high frequency of fire flicker frequency and energy change tendency, combined image processing technology with wavelet. It selects these features as a Support Vector Machine（SVM） input characteristic vector to realize fire recognition. Experimental results show that the proposed method has higher identification accuracy, strong ability to adapt to the environment.

Key words: fire detection, color model, wavelet transform, Support Vector Machine（SVM）