An Adaptive Partial Differential Equation for Noise Removal
Chao Shen, Changxing Li, Pingping Wang · Advances in intelligent systems research/Advances in Intelligent Systems Research · 2013
In order to preserve the image edge feature while removing the noise, an adaptive joint denoising model has been proposed which jointed the second-order and fourth-order partial differential equations (PDE) model.Firstly, improving the traditional fourth-order PDE, separating its diffusion along two orthogonal directions, and then acquiring the separating coefficient fourth-order PDE model.Secondly, this improved fourth-order model combined with the second-order inhomogeneous mean curvature diffusion model, meanwhile, setting the adaptive weights coefficient between second-order and fourth-order models.Finally, the image fidelity term has been added to adaptive joint model.The simulation results show that the adaptive joint model can alleviate blocky and speckle effect effectively, and it has better quantitative results.