Sea-Land Segmentation of Remote Sensing Image Based on Spatial Constraint Model Superpixel Method

Aiming at the problem of sea-land segmentation for maritime target detection, this paper proposes a sea and land segmentation method based on a spatial constraint model merging superpixel regions. Firstly, the remote sensing image is divided into superpixel regions with similar size; secondly, the spatially constrained color texture model is applied to the similarity measurement of adjacent superpixel regions, and finally and the clustering and merging of connected nodes are completed based on the idea of region adjacency graph and complete the division of land and sea. The comparative experiment of the optical remote sensing image database confirms the effectiveness of the proposed method in optical remote sensing image sea-land segmentation.