Application Development Comparative Analysis of AI Based Upscaling and Traditional Upscaling Techniques

Abstract

In this modern era where videos with high resolution are constantly produced and with the current technology advancement making a production of high-quality videos open to everyone and yet the older videos are facing deteriorating image and blur making it hard to recognize the content of the video anymore. In order to prevent that, people invented a method called upscaling where the video can be restored to its former form and thus expand its lifetime and prevent the video becoming a lost media. Since the old video may contain many historical values, it is necessary to preserve it for generations to come. In this paper we will analyze and compare some of the upscaling methods usually used for the video and we will determine which upscaling method offers the best value in restoring and preserving videos. In order to do that we will analyze it using experimental method and display it using a python-based application. The result of this study is that while the AI based upscaling can produce the best image fidelity compared to traditional upscaling method, The AI based method also use an extensive amount of resource making it hard for people to satisfy those demand, and thus making the traditional upscaling method easier to use for many people.