Research On Forest Fire Monitoring Based On Binocular Video Data Fusion

Forest fires pose a serious threat to forest resources and the safety of human life and property.At present,China ’s key forest areas mainly use binocular video to monitor forest fires.In the daytime,visible light video is mainly used to monitor forest fires,while at night,infrared light video is used to monitor forest fires in the early stage of forest fires.However,under the influence of complex environment in forest areas,both monitoring methods have certain limitations.In this paper,Dabie Mountain area of Lu’an City is taken as the research area.Based on the comprehensive analysis of the development status of forest fire monitoring at home and abroad and the theory of image fusion,the research content and technical route are determined.(1)The forest fire monitoring based on visible light uses the background subtraction method to extract the abnormal frame of the dynamic characteristic region of the visible light image,and then the extraction result and the background difference image are processed by threshold segmentation to obtain the forest fire area image.The results show that the use of visible light to monitor forest fires during the day has a good effect on forest fire smoke monitoring,and can monitor the whole process of fire occurrence such as the spread speed and direction of forest fires in real time,which is conducive to forest fire fighting.(2)In the infrared video,the brightness characteristics and dynamic characteristics of the flame high temperature of the forest fire are used to extract the flame feature information by the threshold segmentation method to achieve the monitoring target.The results show that infrared light can accurately monitor the location of fire ignition point in complex forest environment all-weather,and is not affected by light,especially for the initial fire and small area flame under forest,which has an ideal monitoring effect and improves the timeliness of forest fire alarm.(3)Aiming at the characteristics of rich background details in visible images and prominent target areas in infrared images,an image fusion method based on NSCT transform is used for forest fire monitoring.The edge,texture,shape and region features of the fused image obtained by this method are significant.The threshold segmentation is performed on the abnormal frame image and the difference image between the background and the abnormal frame extracted from the fused video by the background subtraction method to monitor the forest fire area.Through comparative analysis,the accuracy of forest fire monitoring is 6.2%and 5.69%higher than that of visible light and infrared light respectively,and the time error is 7.5%and 1.83%lower than that of visible light and infrared light respectively.The use of binocular video data fusion image for forest fire monitoring effectively solves the problem of poor smoke monitoring and recognition effect of forest fires during the night,so as to achieve effective monitoring of forest fires throughout the day and night,and improve the sensitivity of early monitoring of forest fires.The research results provide new ideas and methods for forest fire monitoring and early warning.