Super Resolution Image Reconstruction Using Iterative Regularization Method and Feed –Forward Neural Networks
M. Ganesh Babu, Sudam Sekhar Panda, Homer Benny Bandela · Journal of Physics Conference Series · 2019
Abstract Super resolution is one of the best existing procedures to acquire high resolution image due to its effortlessness and extensive variety of use in numerous fields of science and engineering. There are various methods exist for super resolution but, this work made an effort by combining the iterative regularization method with neural network and obtained a good result comparatively from the previous method because of its high compatible nature with less and user friendly approach. It takes care of the noise in the initial stage and gets a concrete result when neural network is introduced. In addition to the noise it controls the vulnerable parameters and gets a highly super resolution image as compare to the literature.