IDENTIFICATION OF FLOOD DISASTER AREA USING THE DIFFERENTIAL IMAGE WITH THE SEASONAL CHANGES CONSIDERATIONS

Abstract

The satellite images are expected to use for monitoring and identification of the disaster area. A differential image is basic method to distinguish the surface change. It is considered that the differential image is useful to extract the disaster area in which surface condition changed between before and after disaster. However the differential images have not been used efficiently due to seasonal change. This paper proposed a new method to produce a differential image using fine and coarse resolution image. Sensors with coarse spatial resolution can acquire the images with higher temporal frequency. However such coarse image does not have enough spatial resolution to investigate flood disaster area. On the other hand, fine spatial resolution sensor can not acquire the images with higher temporal frequency. The new method simulates the fine resolution image which has the surface reflectance when coarse image was acquired (just before disaster or no seasonal change), using the linear mixture model. New method can exclude the effect of seasonal change and has a good capability to identify only flood disaster area