RECONSTRUCTION OF OPTICAL REMOTE SENSING IMAGES USING SYNTHETIC APERTURE RADAR DATA AND DIFFUSION NEURAL NETWORKS
V.F. Konovalov, V.V. Myasnikov, V.V. Sergeev
Samara University, Samara, Russia
Keywords: remote sensing, digital images, optical satellite images, digital image reconstruction, synthetic aperture radar, cloud removal, neural networks, diffusion
Abstract
This paper suggests a method for reconstruction of optical satellite images using a neural network conditioned on co-registered synthetic aperture radar data. The solution is based on existing neural networks for image reconstruction - DiffCR and Palette, with a modified CBAM block for spatial-channel attention and a modified loss function. The experimental studies conducted show better quantitative and qualitative results in comparison with the baseline models.
|
