This study aims to interpret and analyze images of forest fires and to establish a standard procedure for image processing and interpretation. A forest fire in Victoria, Australia, that occurred in 2009 is used as an example. The extent of the disaster can be analyzed from Formosat-2 images and ALI data. The results show that fire distribution information can be quickly retrieved through scatter plots created by ALI's red and short-wave infrared channels. The burn zones can be rapidly identified from a combination of these wave bands. Moreover, the process of the life of the fire can be deduced through smoke information and changes in the burn zones observed from the images. The maximum likelihood method and K-means method are adopted to rapidly determine the sizes and ranges of the burn zones. The precision obtained by applying this method to images influenced by smoke is 75. 74 %, while that without the influence of smoke is 81. 92 %.
All Science Journal Classification (ASJC) codes
- Water Science and Technology
- Atmospheric Science
- Earth and Planetary Sciences (miscellaneous)