Adaptive rate-distortion optimization for h.264

  • Authors:
  • Kwan-Jung Oh;Yo-Sung Ho

  • Affiliations:
  • Gwangju Institute of Science and Technology (GIST), Gwangju, Korea;Gwangju Institute of Science and Technology (GIST), Gwangju, Korea

  • Venue:
  • PCM'05 Proceedings of the 6th Pacific-Rim conference on Advances in Multimedia Information Processing - Volume Part II
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

In video coding standards, such as MPEG-4 and H.263, one important question is how to determine motion vectors for motion compensation in the INTER mode. Usually the sum of absolute differences (SAD) or the sum of squared differences (SSD) is employed as a matching criterion. Although these criteria are related to the distortion, they do not consider the bit rate appropriately. If we want to consider both the rate and the distortion, a Lagrangian technique targeting for rate-distortion optimization (RDO) is a good alternative. Even if H.264 used the RDO scheme to decide the best macroblock mode among several candidates, H.264 employs only one RDO model for all macroblocks. Since the characteristics of each macroblock is different, each macroblock should have its own RDO model. In this paper, we propose an adaptive rate-distortion optimization algorithm for H.264. We regulate the Lagrangian multiplier according to the picture type and characteristics of each macroblock.