Video Enhancement by Progressive Use of Denoising and Super-Resolution
摘要
Watching a low-resolution video is the troublesome. Now a days this problem is faced due to unavailability of internet. Moreover, there are thousands of old films and videos which are available only in low-quality and low-resolution due to the absence of advanced technology and high-resolution cameras during those time. This is where the concept of video enhancement arrives. Video Enhancement is widely explored yet a challenging domain due to its inability to recreate the content same as original one. The results of using two video improvement techniques, namely denoising and video super-resolution, in sequence is demonstrated in this paper. Two distinct models from denoising and super-resolution are utilized for experimentation. The impacts of four permutations of the said models is tested on the UAV123 dataset, which is one of the largest datasets for high-resolution aerial drone videos.