The Research Of Forest Fire Detection And Early Warning Technology With Double Wave Bands

Forest fire, one of the deadly disasters to the global forestry, causes forest reserves loss and the worldwide environmental pollution at a certain frequency rate each year. There have been numerous researches on pre-warning and suppression measures as soon as forest fire is detected in recent years. The technique of forest fire detection and early warning based on a new model of forest image is proposed in this paper.According to the physical features of forest fire at its early stage, the system of dual band forest fire detection and pre-warning based on infrared and visible light cameras is presented in this study. Smog is detected in the visible band and flames in the infrared band. Two DSP and one FPGA are adopted in the design scheme to realize the ultimate purpose of the forest fire detection and pre-warning system. FPGA is used to, first of all, monitor the collection of visual images by infrared and visible light cameras, and then carry out video image pretreatment. Afterwards, the gathered video images are processed with the assistance of intelligent image processing and pattern reorganization algorithm, based on the characteristics of infrared and visible light images. Finally, a comprehensive evaluation on whether an early-stage forest fire occurs is made based on detection results of infrared and visible light cameras.By conducting a large number of outdoor forest fire detection experiments, the best pre-warning system is identified in this paper. The research shows the system of dual band forest fire detection and pre-warning based on infrared and visible light cameras is flexible in structure and of high generality. It can detect forest fire quickly with high accuracy. The experiment result indicates the present algorithm can be applied to better detect and recognize forest fire at the early stage.