Video Resolution Enhancement

Ajay Pratap Singh; Ashutosh Kumar Shukla; Muskan Pandey

Abstract

In this period of advanced correspondence, video has become a significant aspect of our everyday life. Additionally, video enhancement being a functioning exploration subject in PC vision has gotten more consideration in these ongoing years. Video enhancement manages the accompanying issue: Given an inferior quality info video, how might we make yield video more understood and better". There are essentially two strategies for video enhancement: 1-Spatial - based space video enhancement, 2-Transform - based area video enhancement. Spatial - put together area video enhancement works with respect to pixels. The essential restrictions of this model remember need of giving sufficient heartiness. The fundamental thought behind change-based video enhancement is that to improve the video by controlling the change coefficients. These strategies can't do all the upgrade of the image well overall, and it is hard to mechanize the image enhancement method. Aside from this, actualizing video enhancement utilizing these strategies costs high and is perplexing as well. Perform video enhancement utilizing a solitary image super resolution utilizing four diverse imaging procedures, in particular Super Resolution Convolutional Neural Network (SRCNN), Fast Super Resolution Convolutional Neural Network (FSRCNN), Super Resolution Generative Adversarial Networks (SRGAN), Fast Super Resolution Generative Adversarial Networks (FSRGAN).