MACU-Net for Semantic Segmentation of Fine-Resolution Remotely Sensed Images

TL;DR

This paper introduces MACU-Net, a novel deep learning architecture that enhances semantic segmentation of high-resolution remotely sensed images by integrating multi-scale features and asymmetric convolutions, outperforming existing models.

Contribution

The paper proposes MACU-Net, which combines multi-scale skip connections and asymmetric convolution blocks to improve segmentation accuracy on fine-resolution satellite images.

Findings

MACU-Net outperforms U-Net, U-NetPPL, and U-Net 3+ on remote sensing datasets.

Multi-scale skip connections improve feature alignment across layers.

Asymmetric convolutions enhance feature extraction capabilities.

Abstract

Semantic segmentation of remotely sensed images plays an important role in land resource management, yield estimation, and economic assessment. U-Net, a deep encoder-decoder architecture, has been used frequently for image segmentation with high accuracy. In this Letter, we incorporate multi-scale features generated by different layers of U-Net and design a multi-scale skip connected and asymmetric-convolution-based U-Net (MACU-Net), for segmentation using fine-resolution remotely sensed images. Our design has the following advantages: (1) The multi-scale skip connections combine and realign semantic features contained in both low-level and high-level feature maps; (2) the asymmetric convolution block strengthens the feature representation and feature extraction capability of a standard convolution layer. Experiments conducted on two remotely sensed datasets captured by different satellite sensors demonstrate that the proposed MACU-Net transcends the U-Net, U-NetPPL, U-Net 3+, amongst other benchmark approaches. Code is available at https://github.com/lironui/MACU-Net.