CIF-TO-QCIF VIDEO BITSTREAM DOWN-CONVERSION IN THE DCT DOMAIN

Many advanced video applications require converting a compressed video bitstream from the Common intermediate Format (CIF) to the Quarter Co mmon intermediate Format (QCIF). This task could be performed in the p ixel domain or in the discrete cosine transform (DCT) domain. In this paper, we propose a DCT-domain approach, which requires lower complexi ty than that in the pixel domain. We also propose a scheme for selecti ng a coding mode and a motion vector for a macroblock (MB) in a QCIF v ideo sequence from those of the corresponding CIF video without motion estimation (ME), which is computationally very expensive. Simulation results show that our DCT-domain approach reduces computational comple xity by about 40% compared to the pixel-domain approach while providin g comparable picture quality with about 0.4 dB degradation.