Distortion identification technique in video stabilization

Abstract

For non-stationary displacement sequence of the real-world image sequence, this paper presents a distortion identification technique based on Hilbert Huang Transform (HHT) to identify the distortion model and distortion frequency of the displacement sequence. Experiment results show that this technique can identify the distortion model and distortion frequency of the displacement sequence accurately and quickly. Based on these identification results, we can realize the video stabilization effectively.