Prediction of Pair Trawl fishing Activity using Sar and AIS Data

Abstract

In spite of several research works on ship detection from satellite images success in the detection of small fishing vessels is still remarkably low. Therefore, a method is proposed to detect small ships from Sentinel-1 images for the purpose of further identification of trawl fishing ships and finally to predict the pair trawl fishing. For this purpose, an image contrast enhancement was applied to remove background and keep only the objects in images. We applied this method to one Sentinel-1 GRDH image covering the Socheongcho area in the Yellow Sea and achieved 91% accuracy in small trawl ship detection. Then automatic identification system data was used to identify the pair trawl fishing. Thus, it is expected that application of the proposed methodology could enhance the marine surveillance.