Flood extent mapping method by means of mathematical morphology using L band SAR data: Discrimination of damaged areas from ordinary inundated areas in paddy fields
2007
Yamada, Y.(National Inst. of Rural Engineering, Tsukuba, Ibaraki (Japan)) | Tingsanchali, T. | Khao-Uppatum, V. | Ishizaka, K. | Nakamura, Y.
Sometimes agricultural fields suffer from flood damages for a long period along a large river in the tropical regions. It is very important to detect the flood extent in order to assess flood damage to agriculture. L-band SAR images are effective in detection of the damaged areas at flood disaster with the advantage of the character of penetrating cloud cover. The authors tried to detect the flood-inundated and damaged areas using SAR data by the following approach at the Central Plain of the Chao Praya basin in Thailand. Mathematical morphology operations and directional filters provide image feature measurements. Terrain feature recognition for SAR imagery combines both target's physical property and morphological image processing. This method should be adopted in order to solve the problem how to distinguish paddy fields from flood inundated areas instead of using the conventional SAR signal intensity change method. For example, if the Levee of paddy fields is above the inundated water surface, the paddy fields should be supposed almost without damages in the consolidated farm lands. If the region of the low backscattering intensity level occupy large area, that can be most likely calm and large water surface or flood-inundated areas. These calculations require only a snapshot of SAR image. This image processing procedure has advantage over the flood disaster monitoring using SAR satellite data.
显示更多 [+] 显示较少 [-]