PROJECT TITLE :

Single Image Super-Resolution Based on Wiener Filter in Similarity Domain - 2018

ABSTRACT:

Single image super-resolution (SISR) is an sick-posed problem aiming at estimating a plausible high-resolution (HR) image from one low-resolution image. Current state-of-the-art SISR strategies are patch-primarily based. They use either external data or internal self-similarity to be told a prior for an HR image. External data-based strategies utilize a giant range of patches from the training information, while self-similarity-based approaches leverage one or more similar patches from the input image. During this Project, we tend to propose a self-similarity-based mostly approach that's in a position to use massive groups of similar patches extracted from the input image to solve the SISR problem. We introduce a novel previous leading to the collaborative filtering of patch groups during a 1D similarity domain and couple it with an iterative back-projection framework. The performance of the proposed algorithm is evaluated on a number of SISR benchmark information sets. Without using any external data, the proposed approach outperforms the present non-convolutional neural network-based ways on the tested information sets for varied scaling factors. On certain information sets, the gain is over 1 dB, when compared with the recent method A+. For prime sampling rate (x4), the proposed method performs similarly to terribly recent state-of-the-art deep convolutional network-based approaches.