Transformer-Based SAR Image Despeckling

• Published in: IGARSS 2022 - 2022 IEEE International Geoscience and Remote Sensing Symposium pp. 751 - 754
• Main Authors: Perera, Malsha V., Bandara, Wele Gedara Chaminda, Valanarasu, Jeya Maria Jose, Patel, Vishal M.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 17-07-2022
• Subjects: Apertures; Convolutional codes; denoising; Geoscience and remote sensing; Network architecture; Radar polarimetry; Speckle; Synthetic Aperture Radar; Transformers
• ISSN: 2153-7003
• DOI: 10.1109/IGARSS46834.2022.9884596

Synthetic Aperture Radar (SAR) images are usually degraded by a multiplicative noise known as speckle which makes processing and interpretation of SAR images difficult. In this paper, we introduce a transformer-based network for SAR image despeckling. The proposed despeckling network comprises of a transformer-based encoder which allows the network to learn global dependencies between different image regions - aiding in better despeckling. The network is trained end-to-end with synthetically generated speckled images using a composite loss function. Experiments show that the proposed method achieves significant improvements over traditional and convolutional neural network-based despeckling methods on both synthetic and real SAR images. Our code is available at: https://github.com/malshaV/sar_transformer