Small- to Medium-Sized Stream Waterbody Extraction Using Sentinel-1 Images Based on Shortest Plane Distance

In this study, a method for extracting the waterbody of small- to medium-sized streams is verified using SAR (Synthetic Aperture Radar) images based on the shortest plane distance (DST). The areas investigated are the Hantangang River and Jungnangcheon stream. To verify the waterbody extraction method, SAR satellite image data (Sentinel-1) and PlanetScope, which is a high-resolution optical satellite, are used simultaneously. After preprocessing the Sentinel-1 images, their brightness distribution is equalized via histogram matching. To realize an efficient stream extraction, a weight, which is the DST from the stream centerline, is added. After combining this weight value and Sentinel-1's VH and VV polarizations, the optimal parameter value is obtained using the k-means method. This value allows the waterbody to be extracted from Sentinel-1 images with maximum accuracy, depending on the resolution limit. Based on the correlation among the VV, VH, and DST, the waterbody extraction can be calculated using an elliptic equation. Results show that the average accuracy is 0.45-0.75, and the average Kappa coefficient is 0.60-0.85. This study shows that the waterbody area can be estimated using the DST. Additionally, the proposed method extracts the waterbody more simply and rapidly than the existing waterbody extraction method.