SDRNET: Stacked Deep Residual Network for Accurate Semantic Segmentation of Fine-Resolution Remotely Sensed Images

TL;DR

This paper introduces SDRNet, a stacked deep residual network that improves semantic segmentation of high-resolution remotely sensed images by capturing multi-scale features and global context, leading to more accurate land cover maps.

Contribution

The paper proposes a novel stacked encoder-decoder architecture with dilated residual blocks to enhance feature extraction and spatial detail preservation in FRRS image segmentation.

Findings

SDRNet outperforms existing methods on ISPRS Vaihingen and Potsdam datasets.

The model effectively captures multi-scale and global contextual features.

Results show improved boundary accuracy and class discrimination.

Abstract

Land cover maps generated from semantic segmentation of high-resolution remotely sensed images have drawn mucon in the photogrammetry and remote sensing research community. Currently, massive fine-resolution remotely sensed (FRRS) images acquired by improving sensing and imaging technologies become available. However, accurate semantic segmentation of such FRRS images is greatly affected by substantial class disparities, the invisibility of key ground objects due to occlusion, and object size variation. Despite the extraordinary potential in deep convolutional neural networks (DCNNs) in image feature learning and representation, extracting sufficient features from FRRS images for accurate semantic segmentation is still challenging. These challenges demand the deep learning models to learn robust features and generate sufficient feature descriptors. Specifically, learning multi-contextual features to guarantee adequate coverage of varied object sizes from the ground scene and harnessing global-local contexts to overcome class disparities challenge even profound networks. Deeper networks significantly lose spatial details due to gradual downsampling processes resulting in poor segmentation results and coarse boundaries. This article presents a stacked deep residual network (SDRNet) for semantic segmentation from FRRS images. The proposed framework utilizes two stacked encoder-decoder networks to harness long-range semantics yet preserve spatial information and dilated residual blocks (DRB) between each encoder and decoder network to capture sufficient global dependencies thus improving segmentation performance. Our experimental results obtained using the ISPRS Vaihingen and Potsdam datasets demonstrate that the SDRNet performs effectively and competitively against current DCNNs in semantic segmentation.