Abstract
Semantic segmentation for very-high-resolution remote sensing images has been a research hotspot in the field of remote sensing image analysis. However, most existing methods still suffer from a challenge that object boundaries cannot be finely recovered. To tackle the problem, we develop a dual-stream network based on the U-Net architecture, Instead of the traditional skip connections, a boundary attention module is proposed to introduce the boundary information from the EDN module to the SSN module. Experiments on ISPRS Potsdam and Vaihingen datasets show the effectiveness of the proposed network, especially in man-made objects with distinct boundaries.
Highlights
I N THE field of remote sensing, semantic segmentation of very-high-resolution (VHR) remote sensing images is of significance to many applications, such as land cover mapping, urban change detection, and environmental monitoring [1]–[3]
We first present the overview of the proposed dual-network for semantic segmentation of VHR remote sensing images
It is demonstrated that our method outperforms other methods in terms of mean F1-score and overall accuracy (OA)
Summary
I N THE field of remote sensing, semantic segmentation of very-high-resolution (VHR) remote sensing images is of significance to many applications, such as land cover mapping, urban change detection, and environmental monitoring [1]–[3]. Automatic semantic segmentation for remote sensing image has become a fundamental problem for a long time. As the outstanding performance of convolutional neural networks (CNNs) in computer vision, recent researches have proven CNNs very successful tools for VHR remote sensing images semantic segmentation [4]. FCN [7] is a milestone for semantic segmentation and many FCNbased methods [8]–[10] have been developed.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have