Data partitioning for parallel implementation of real-time video processing systems

The main reason for using data partitioning for a video processing system is the increase in performance it offers. This paper presents a comparison of the buffering requirements for different data partitioning methods used in video processing systems. The analysis shows that the buffer storage required for the implementation of video systems is highly dependent on the partitioning strategy. The results indicate that partitioning the tasks vertically is by far the most efficient method when considering buffer sizes.