Abstract
Semantic segmentation for remote sensing images (RSIs) is widely applied in geological surveys, urban resources management, and disaster monitoring. Recent solutions on remote sensing segmentation tasks are generally addressed by CNN-based models and transformer-based models. In particular, transformer-based architecture generally struggles with two main problems: a high computation load and inaccurate edge classification. Therefore, to overcome these problems, we propose a novel transformer model to realize lightweight edge classification. First, based on a Swin transformer backbone, a pure Efficient transformer with mlphead is proposed to accelerate the inference speed. Moreover, explicit and implicit edge enhancement methods are proposed to cope with object edge problems. The experimental results evaluated on the Potsdam and Vaihingen datasets present that the proposed approach significantly improved the final accuracy, achieving a trade-off between computational complexity (Flops) and accuracy (Efficient-L obtaining 3.23% mIoU improvement on Vaihingen and 2.46% mIoU improvement on Potsdam compared with HRCNet_W48). As a result, it is believed that the proposed Efficient transformer will have an advantage in dealing with remote sensing image segmentation problems.
Highlights
Academic Editors: Mercedes E.Paoletti and Juan M
This paper makes a thorough study of applying a Swin transformer on remote sensing segmentation and proposes an Efficient transformer backbone, pure transformer segmentation head named mlphead, explicit edge enhancement and implicit edge enhancement methods to cope with three main shortcomings of Swin transformer: high computation load, low inference speed, and low ability to extract edge details
The experimental results are arranged as following parts: the study of Swin transformer segmentation model, ablation experiments regarding to our proposed architectures, and comparison with SOTA models on the Potsdam and Vaihingen datasets
Summary
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Remote sensing (RS) is capable of collecting information from objects with no physical contact through satellite and aerial-based platforms and is widely applied in geological survey, urban resources management, and disaster monitoring [1,2]. To achieve a high performance in the aforementioned applications, an automatic and fast detection method for ground objects with high accuracy in a robust way is desirable. The deep learning method, a pioneering method in remote sensing area that develops rapidly, with high accuracy, robustness, and fast inference speed, is widely used in object detection and classification tasks
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
