Extraction of aquaculture ponds based on U2-net using remote sensing images in the Liuheng island, China

Abstract

As the third largest island in Zhoushan, China, Liuheng island is one of the important pillars of Zhoushan aquaculture. Accurately extracting information such as the distribution and area of aquaculture ponds is of great significance to the future planning and management of aquaculture ponds. In this study, U2-net was used to extract the characteristics of coastal aquaculture ponds in Liuheng island, China. The results show that this method can reduce the influence of the same-spectrum foreign body phenomenon and effectively improve accuracy. During the study period of 2020, in terms of time distribution, the area of aquaculture ponds in Liuheng island fluctuated from 5.8 km2 to 8.6 km2 with the seasons. In terms of spatial distribution, coastal aquaculture ponds of Liuheng island are mainly distributed in the southern part of the island (5.39 km2). This study could help producers manage coastal aquaculture more scientifically.