Recently, there has been greatly increased interest in integrating multimedia content into computer applications. One of the most challenging aspects of that task has been the effective use of video. The accurate representation of video information takes a tremendous amount of data. This amount of data is so great that it must be compressed to be effectively communicated with current computer hardware.
One particularly effective compression scheme is the international standard Motion Pictures Expert Group (MPEG) suite of algorithms. Although MPEG encoding allows for the compression of both audio and video of a range of qualities, this project concentrates on MPEG-1 video compression because it is specifically aimed at producing data rates optimized for computer communication and display.
MPEG-1 video compression is a lossy process. In order to reduce the data rate to an acceptable level, some of the data is omitted and can not be recovered. To achieve this compression, the algorithm reduces both temporal redundancy and spatial redundancy in the encoded data. Spatial redundancy refers to the fact that, at any instance in time, the image in a very small area has a high probability of being similar. Temporal redundancy means that in a short amount of time, a small area is likely to stay in the same place or move as a unit to a new location.
MPEG-1 video compression's effectiveness comes at the cost of speed. Until recently, real-time compression required thousands or tens of thousands of dollars worth of dedicated hardware. Now, inexpensive encoders can be found that sell for less than $1000. However, these prices still put these products out of the reach of most individuals. One of the important properties of this compression technique is that it is highly parallel. The approach of this project is to exploit that characteristic by performing the compression in software running on a group of generic, inexpensive computers working together on a single video input. This will provide a cost/performance advantage and will make the process available to a greater range of users.