N-Net: a remote sensing image segmentation network with edge enhancement induction module

Aiming at the issue of rough edges and misidentification in remote sensing image segmentation using U-Net, this paper designs a deep learning image segmentation network (N-Net) which integrates an edge enhancement induction module. After training and testing on the WHU dataset, N-Net achieves better building edge recognition and extraction capabilities compared to U-Net, with minimal increase in computational complexity.