Hyperspectral image (HSI) mixed denoising is a challenging task in the fields of remote sensing, environmental monitoring, mineral exploration, and so on. A crucial difficulty is to acquire clean restoration… Click to show full abstract
Hyperspectral image (HSI) mixed denoising is a challenging task in the fields of remote sensing, environmental monitoring, mineral exploration, and so on. A crucial difficulty is to acquire clean restoration from HSIs that encounter Gaussian noise, impulse noise, strip noise, and deadlines. In the previous works, combining global information and nonlocal information is a popular way to learn the comprehensive characteristics of the clean HSI. However, the advantages of 2-D spatial structure similarity and spectral low-rankness may not be fully exploited at the same time in global prior learning. The iterative update between global restoration and nonlocal restoration may cause high time consumption and a certain loss of information. To address these issues, we propose a three-stage mixed denoising model based on novel hybrid low-rank tensor approximation and global-guided-nonlocal prior mechanism (HLTA-GN). First, to learn a good global prior, a hybrid low-rank tensor approximation incorporated with a useful nonconvex tensor rank estimation is presented to balance 2-D spatial similarity and spectral low-rankness. Second, to learn a high-quality nonlocal prior, a global-guided-nonlocal prior mechanism is proposed to help nonlocal restoration suppress the residual noise. At the same time, a regularized sequential low-rank tensor approximation is proposed to enhance the robustness of nonlocal optimization to noisy patch groups. Third, a weighted fusion of global prior and nonlocal prior helps to further balance global denoising and patch processing. An efficient learning algorithm is provided to solve HLTA-GN. Abundant experiments are conducted on various HSIs with several scenarios. The experimental results demonstrate the superiority of HLTA-GN.
               
Click one of the above tabs to view related content.