Advances in video compression

The market for digital video is growing, with compression as a core enabling technology. Standardized methods as motion-compensated DCT, which have reached a level of maturity over decades, are dominating today's products. During the past years, another wave of innovation has occurred in video compression research and development, which leads to the assumption that it is still far from reaching its final bounds. Key factors are further improvements in motion compensation, better understanding of spatio-temporal coherences over shorter and longer distances, and more advanced encoding methods, as e.g. implemented in the new Advanced Video Coding (AVC) standard. While the adoption of successful research trends into standardization and products seems to occur almost seamlessly, new trends are already showing up which may lead to more paradigm shifts in the video compression arena. In general, the interrelationship between transmission networks and compression technology bears many problems yet to be solved. Efficient scalable representation of video becomes more interesting now, providing flexible multi-dimensional resolution adaptation, to support various network and terminal capabilities and better error robustness. This has finally become possible by the advent of open-loop temporal compression methods, denoted as motion-compensated temporal filtering (MCTF), which are presently investigated for standardization. More improvements seem to be possible in the fields of motion compensation and texture representation, but also new application domains are emerging. 3D and multi-view video seems will become more realistically used in applications, due to the availability of appropriate displays. Mobile and sensor networks are raising a future demand for low-complexity encoders. The talk will analyse some of these trends and investigate perspectives and potential developments.