研究目的
Investigating flood area detection using ALOS-2 PALSAR-2 data acquired during the 2015 heavy rainfall disaster in the Kanto and Tohoku areas, Japan.
研究成果
The study demonstrated the feasibility of using PALSAR-2 data for rapid flood monitoring, with lower off-nadir angles (around 30 degrees or less) being effective in flood area detection. Polarimetric analysis and speckle filtering were found to improve accuracy.
研究不足
The method is not effective for flooding in built-up areas and agricultural areas, which will be addressed in future work.
1：Experimental Design and Method Selection:
The study used thresholding of PALSAR-2 amplitude images for flood detection, comparing different speckle filters and off-nadir angles.
2：Sample Selection and Data Sources:
PALSAR-2 data covering the city of Joso, Japan, during the 2015 heavy rainfall disaster, and inundation maps based on aerial photographs as reference data.
3：List of Experimental Equipment and Materials:
ALOS-2 PALSAR-2 data, aerial optical photographs for reference.
4：Experimental Procedures and Operational Workflow:
Speckle filtering, geometric correction, flood detection by thresholding back-scattering, and accuracy evaluation against reference data.
5：Data Analysis Methods:
Comparison of flood detection accuracies using overall accuracy (OVA) and kappa coefficient (κ).
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
