Abstract

Image super-resolution is a technique that seeks to improve the sharpness and fidelity of images with low resolutions. This process involves the reconstruction of high-resolution images from their lower-resolution equivalents, thereby enhancing the visual intricacies and overall clarity of the image. Profound comprehension of the principles governing image and video superenhancement holds significant importance in the formulation of efficient algorithms and methodologies across diverse domains. Super-resolution is a major task in various imaging applications like medical, satellite and agriculture particularly recovering structures of interest. By considering spatial relationships, these methods aim to capture fine details present in underlying structures. Many researchers have done survey on various methods used for super resolution. The objective of this review is to offer a thorough examination of the diverse approaches employed in enhancing the resolution of images, different datasets used, and scaling factors for images used by the researchers for various applications. Finally the paper reviews the various models of Generative Adversarial Neural Networks which proves to be state of the art method for image super resolution.