PROJECT TITLE :

Patch-Based Video Denoising With Optical Flow Estimation

ABSTRACT:

A unique image sequence denoising algorithm is presented. The proposed approach takes advantage of the self-similarity and redundancy of adjacent frames. The algorithm is inspired by fusion algorithms, and as the amount of frames increases, it tends to a pure temporal average. The employment of motion compensation by regularized optical flow methods permits robust patch comparison during a spatiotemporal volume. The utilization of Principal Component Analysis ensures the correct preservation of fine texture and details. An intensive comparison with the state-of-the-art ways illustrates the superior performance of the proposed approach, with improved texture and detail reconstruction.