APPLICATION FOR SHIP DETECTION USING SKEWNESS AND KURTOSIS

SAR remote sensing has emerged as an efficient tool for monitoring vast ocean areas due to its wide swath coverage and operational flexibility. In this context, many applications have been developed for ocean traffic monitoring based on conventional ship detection techniques. On the other hand, the low temporal resolution of satellite-based SAR for ocean monitoring has been a problem. Nowadays, this issue has been addressed by building constellations of satellite-based SAR systems. However, as the number of satellites increases the volume of observation data also increases. It will be valuable to quickly detect the distribution of ships from this vast amount of observation data. In this paper, we propose a coarse-to-fine approach to quickly detect the area with ships. We adopted the skewness and kurtosis of the distribution of backscatter from sea surface for coarse detection. We evaluated the effectiveness of an application by X-band C-band and Lband SAR data.