Real time implementation of rate-distortion optimized coding mode selection for H.263 video coders

This paper proposes a new method for real time implementation of rate-distortion optimized coding mode selection which can be efficiently applied to H.263-compatible video codecs and other codecs of similar type. We use our previously proposed normalized rate-distortion model to efficiently compute the rate and the distortion when encoding motion-compensated prediction error signals, instead of actually performing DCT, quantization and entropy-encoding. We also propose a fast algorithm to find sub-optimal values of coding parameters such as the quantization parameter and the Lagrangian multiplier, λ, for the trellis search. Experiments show that the proposed scheme provides very good rate control as well as good picture quality, especially when applied to very low bit rate video coding