Towards a partial differential equation remote sensing image method based on adaptive degradation diffusion parameter

For the anisotropy diffusion feature, Partial Differential Equation (PDE) methods keep edge detail characters well in case of denoising, thus being widely applied in remote sense image denoising, smoothing, filtering and reconstruction. A PDE remote sensing image denoising method based on Adaptive Degradation Diffusion Parameter (ADDP) was proposed in the paper to deal with fuzzy detail problem caused by increasing iteration number. The PDE denoising method with ADDP enlarged diffusion size in the plat region of remote sensing image without affecting the remote sensing image edge, thus avoiding loss of remote sensing image detail and intersections caused by Gaussian convolution smoothing in the PDE filtering model based on curvature-based movement (CM) and image denoising model based on total variation (TV). In the region where gradation value changes little, the method executed isotropic diffusion to remove isolated noise. The upwind scheme was applied for model numerical realization. Experimental in remote sensing image denoising results proved its feasibility and effectiveness.