Motion-Compensated Temporal Filtering for Critically-Sampled Wavelet-Encoded Images

TL;DR

This paper introduces a novel motion estimation and compensation method directly on wavelet coefficients, improving low-bitrate video coding quality without requiring shift-invariance or interpolation.

Contribution

It presents a new in-band motion compensation technique that operates directly on DWT coefficients, avoiding redundancy and upsampling, with exact subpixel accuracy.

Findings

Achieves high video quality at very low bitrates

Demonstrates effective motion compensation on wavelet coefficients

Improves low-bitrate video coding performance

Abstract

We propose a novel motion estimation/compensation (ME/MC) method for wavelet-based (in-band) motion compensated temporal filtering (MCTF), with application to low-bitrate video coding. Unlike the conventional in-band MCTF algorithms, which require redundancy to overcome the shift-variance problem of critically sampled (complete) discrete wavelet transforms (DWT), we perform ME/MC steps directly on DWT coefficients by avoiding the need of shift-invariance. We omit upsampling, the inverse-DWT (IDWT), and the calculation of redundant DWT coefficients, while achieving arbitrary subpixel accuracy without interpolation, and high video quality even at very low-bitrates, by deriving the exact relationships between DWT subbands of input image sequences. Experimental results demonstrate the accuracy of the proposed method, confirming that our model for ME/MC effectively improves video coding quality.