Abstract:

The reliable and up-to-date information on the water resources is increasingly important, not only for the management scopes but also for farmers and urban planners. Microwave active remote sensing technology has the all-time and all-whether advantage for the earth observation. It is not affected by night and fog and has ability to penetrate through the cloud during the flood period. A new generation of synthetic aperture radar sensors, called Sentinel-1A, has a much improved spatial, temporal, and radiometric resolution. The Sentinel1A, with four exclusive imaging modes of operations: Interferometric Wide Swath, Extra Wide Swath, Strip Map, and Wave, provides the obviousness for water and land boundary in the SAR images due to low backscatter value in the pixels of the water surface. As the results, analyzing images by using image thresholding techniques is often used to extract water surface from SAR images. In this study, an automatic method is presented to identify water body areas via using minimum cross entropy, which is also known as Li's minimum cross entropy method from Sentinel-1A IW images. The evaluation of the proposed method shows a high accuracy with 99.5% of the user accuracy, 98.8 % of the producer accuracy, and 97.9 % of the overall accuracy