Parallel Advanced Video Coding: Motion Estimation on Multi-cores

Abstract

The new Advanced Video Coding (AVC) standards further exploit temporal correlation
between images on a sequence by considering multiple reference
frames and variable block sizes. It increases the compression rate
for a given video quality at the cost of a significant increase in
the computational load. Specialized hardware processors have been
proposed to perform real time motion estimation on AVC, but the
non-recurring engineering cost of these solutions is too high. This
paper proposes a parallel algorithm that exploits the capacity of
the current multi-core processors to implement real time motion
estimation for AVC. In particular, exploiting the computational
capacity and the fast memory system of the heterogeneous multi-core
CELL processor, the synergetic processors accelerate the motion
estimation while the main processor executes in parallel the other
components of the AVC system. Experimental results show that
motion estimation can be performed in less than 50ms per frame, for
CIF video format, with up to 5 reference frames and variable block
sizes, by programming the CELL with the proposed parallel algorithm.
In addition, the scalability of the proposed solution is proven
regarding the video sequence resolution, the number of cores and
reference frames used.