A Successive Variational Model for Multiplicative Noise and Blur Removal
Suman Kumar Maji, Ramesh Kumar Thakur · 2022 IEEE 19th India Council International Conference (INDICON) · 2022
In this paper, we propose a novel variational formulation for blur and speckle noise removal from images. The proposed model operates on the concept of applying blur removal (deblurring) and noise removal successively on the observed sample. Such successive application of despeckling and deblurring, in a Bayesian framework, is being realized by reformulating the image formation model that we also propose. The proposed optimization problem is then minimized using Euler-Lagrange formulation. Visual results suggest the potential of the proposed method in restoring images corrupted by both blur and speckle noise. Further qualitative and quantitative analysis validate its superiority over the state-of-the-art methods.