The Research of Real-Time Forest Fire Alarm Algorithm Based on Video

Forest fire has brought incalculable economic losses due to its strong abruptness and huge destructiveness. Forest real-time monitoring and fire alarm can effectively reduce the loss of life and property, but the traditional detection devices are not widely used for its limited detection range. This paper proposes a real-time forest fire alarm algorithms based on video. Extracting the moving region on the basis of background modeling, and then we use the mixed color space feature analysis to detect the flame, and use adaptive threshold segmentation and color moments analysis to detect smoke. The experiments has shown that our algorithms have both a higher detection rate and a lower false alarm rate, and the average processing speed reaches 40 ms/frame, and the detection of flame and smoke under different environments and lighting conditions can also work accurately and quickly.