Low complexity H.264/AVC spatial resolution transcoding in the transform domain

In this paper, we propose a novel efficient downsizing video transcoder for H.264/AVC low bit-rate compressed images. The algorithm accounts for new coding tools introduced by the H.264/AVC standard and applies directly on the transformed coefficients, thus reducing greatly the computational complexity of the transcoding algorithm. In a first approach, we restrict ourselves to 4×4 blocks intra-coded frames. Simulation results are presented to illustrate the performances of the proposed method. In particular, we show that the computational complexity is greatly reduced compared to the optimal full decode-full recode approach, while giving similar reconstructed video quality at low bit rates.