A Multiscale Dual Attention Network for the Automatic Classification of Polar Sea Ice and Open Water Based on Sentinel-1 SAR Images

作     者：Zhang, Zheng Deng, Guangbo Luo, Chuyao Li, Xutao Ye, Yunming Xian, Di 

作者机构：Harbin Inst Technol Dept Comp Sci Shenzhen 518055 Peoples R China Harbin Inst Technol Shenzhen Key Lab Internet Informat Collaborat Shenzhen 518055 Peoples R China China Meteorol Adm Natl Satellite Meteorol Ctr Beijing 100081 Peoples R China 

出 版 物：《IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING》 (IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens.)

年 卷 期：2024年第17卷

页      面：5500-5516页

核心收录：

学科分类：0808[工学-电气工程] 1002[医学-临床医学] 08[工学] 0705[理学-地理学] 0816[工学-测绘科学与技术] 

基　　金：Shenzhen Science and Technology Program  China 

主　　题：Sea ice Microwave radiometry Feature extraction Radar polarimetry Transformers Synthetic aperture radar Microwave theory and techniques Climate change Deep learning Noise measurement Learning systems sea ice classification synthetic aperture radar (SAR) 

摘      要：Automatic classification of sea ice and open water plays a vital role in climate change research, polar shipping, and other applications. Many deep-learning-based methods are proposed to automatically classify sea ice and open water to address this issue. Even though these methods have achieved remarkable success, the noise phenomenon in synthetic aperture radar (SAR) images still causes considerable limitations in the model performance. Meanwhile, these existing methods ignore multiscale global information from large-scale SAR images, which tends to produce misclassification. In this article, we propose a novel multiscale dual attention network (MSDA-Net) for the task. To tackle the first drawback, we introduce the information of relative position and high-pass filtering as two extra channels to reduce the noisy effects. Moreover, we propose a patch dual attention mechanism and embed it into the ConvNeXt blocks to capture the multichannel and spatial features. To address the second problem, we propose a multiscale spatial attention module to capture multiscale global spatial information. The experiments show that the proposed method significantly outperforms state-of-the-art methods. In addition, comprehensive case studies are conducted, which verify the effectiveness of MSDA-Net in different SAR scenes.