ABSTRACT
Flood disaster is extracted and mapped based on SAR data and optical data. The flood disaster monitoring method based on image classification has higher recognition accuracy than other methods in the completely inundated area and vegetation inundated area and also identifies a large number of other land types, which is greatly improved when compared with the three commonly used methods. Based on Sentinel-1 data, two types of inundated area extraction models for open water and vegetation area were constructed and applied to flood monitoring in the lower reaches of the Ilowadi River in Myanmar, and good results were obtained in the extraction of the cultivated land affected area. The flood disaster monitoring model takes image classification and change detection as the core and SAR as the main data source. Many scholars also use the vegetation index to study flood extraction.
