Prototype for Integration of colorization and super resolution model using machine learning
Anirudh Lodh, Utkarsh Saxena, Ajmal Khan, K.A. Patel Himey Atulkumar, S. Sountharranjan · 2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA) · 2021
Colorization is the process of converting black and white images to color digitally. On the other hand, super resolution is the process of enhancing the quality of low-resolution pictures by converting them into high resolution, while maintaining the original details of the picture. The idea behind our implementation is to combine these two machine learning models. The combined model will take in old, blurry, hazed images as input and produce colorized, detailed and high-resolution image as an output. This research work has utilized various machine learning and computer vision techniques to accomplish this task. For colorization, classification based deep learning model was utilized. For super resolution, Post-up-sampling convolutional neural networks have been used.