New Fast Detection Method of Forest Fire Monitoring and Application Based on FY-1D/MVISR Data

In this paper, the authors proposed a new work and process flow algorithm about remote sensing image data to forest fire identification and monitoring, which was greatly different with the traditional approaches. Therefore, a more useful context method was used to detect forest fire spots, banes on statistic rationale, meanwhile the cloud-contaminated pixels were rejected (if any) and the misjudged fire spots excluded by examination of NDVIs before and after forest fires occurred; moreover, a simpler and more resultful approach was used to earth-locate forest fire spots, and a forest-fire sub-pixel area situation was taken into account so as to evaluate forest-fire area and enhance accuracy of evaluated forest fire areas. Subsequently, a FFDM prototype system was designed and developed and later tested using the typical FY-1D/MVISR data, performance and running efficiency of which were practically greater than the normal mainstream software systems (e.g. ENVI) in this domain. Key-Words: FY-1D/MVISR, Radiance, vegetation index, geo-referance, workflow New Fast Detection Method Of Forest Fire Monitoring And Application Based On FY-1D/MVISR Data 621