Research On The Realization And Application Of Fire Point Extraction Algorithm Based On Ground Satellite Receiving Station Data

Forest fires are sudden and destructive natural disasters,which have irreversible impacts on ecological environment.Therefore,it is of great significance to master forest fires in a timely manner.Traditional monitoring methods cost lots of manpower and material resources,and are vulnerable to the interference of weather and geographical environment,which often fail to meet the actual needs of detecting forest fires.In order to meet the demand of monitoring fires in north China,the algorithm based on adaptive threshold was proposed to improve the detection accuracy of forest fires.To analysis the accuracy of fire detection algorithms,seven forest fires in north China since 2017 were selected to get the quantity and spatial distribution of fires.MODIS data was used to detect fires.Take the medium and spatial resolution data as the comparison standard,the accuracy comparison of algorithms were carried out.By analyzing the quantity and spatial distribution of fire hotspots,the accuracy evaluation and model optimization of the algorithms were carried out.The results indicated that MODIS fire hotspots algorithm and fire hotspots recognition algorithm based on variance between classes could not detect all the forest fires.Among the data of medium and high spatial resolution,the fire algorithm based on adaptive threshold was significantly better than the other four algorithms.At the same time,the omission rate and false detection rate of fire algorithm based on adaptive threshold were lowest among the other algorithms.When filtering potential fire pixels,the improved algorithm introduce the idea of adaptive thresholds,which can be adjusted and improved for more accurate regional fire detection.In general,combined with the improved algorithm,data from satellite ground receiving station can apply to real-time supervision of fires.