A New Approach to ATV Video - Decoding
A key technology of digital television is the digital video data compression algorithm known as MPEG-2; the international standard created by the Motion Pictures Engineering Group. The MPEG-2 digital video compression standard has become the universal standard in ATV and other multi-media systems.
HOW IT WORKS:
MPEG-2 uses a motion compensated discrete cosign transform (DCT) algorithm. The DCT exploits spatial redundancy, and motion compensation exploits temporal redundancy. The video encoder of the ATV system provides MPEG-2 syntax compatible bit streams which supports both predictive picture (P-picture) and bi-directional predictive picture (B-picture). It also supports wide motion-estimation ranges, field and frame motion vectors, and adaptive field and frame DCT coding to improve the compression efficiency of interlaced formats. However; the traditional MPEG-2 decoder stores two reconstructed frames in the spatial domain. This has presented significant cost considerations in ATV decoder design.
OUR APPROACH:
MERL Murray Hill has developed a more efficient memory saving architecture for the ATV decoder. With MERL Murray Hill's new architecture, the motion compensation is performed in the frequency domain (MC-DCT) instead of the spatial domain. In this way, the inverse DCT is of the motion compensation loop. This makes it possible for the anchor frames to be saved in the compressed bits domain instead of the spatial domain. The result is significant savings in the memory requirements, even with the addition of several extra devices in the decoding loop. The simulation results with the new decoder architecture have shown that the reconstructed image quality is equal, if not superior, to that of the conventional decoder.
Contact: Huifang Sun
Technology Area: Advanced Digital Television
Modification Date: August 2, 2004
