基于视频的火焰检测技术研究 Flame-Detection Technique Based on Video Processing

综合分析了火焰的颜色、闪烁频率及其面积增长性等静态特征和动态特征,设计了一种基于视频的火焰检测自动预警算法. 利用混合高斯模型有效地表示了火焰颜色及颜色的多态性. 实验表明该算法具有误报警低、识别率高、实时性强等优点,是火灾自动报警的一种可行、科学的解决方案,具有广泛的应用前景,该研究在国内外还处于发展阶段. The paper analyzes static characteristics and dynamic ones of flame such as flame s colors、flickering frequency and augmentation,puts forward a new flame-detection algorithm that could automatically give an alarm on video processing.Flame s color and its polymorphism were described effectively by mixture Gauss model.The experimental results show that the flame-detection algorithm has many advantages such as lower false alarm rate,higher recognition rate and it can be used in many real-time control environme...