Rate-distortion motion compensated noise filtering for low bit rate video coding

Noise in video coding applications degrades the image quality and prevents effective compression. In this paper, we define a rate-distortion optimization framework to adapt the response of a motion compensated (MC) temporal noise filter for MC hybrid video coding schemes. We propose a solution based on the Lagrange multiplier method to determine the optimal control parameter set of the filter, and we describe an effective technique to estimate the a priori unknown rate constraint. The results demonstrate that the proposed method leads to near-optimal compression ratio and satisfactory distortion performance.