A real-time implementation of the efficient 2-D discrete cosine transform

The authors introduce a real-time implementation of the 2-D discrete cosine transform (DCT). The implementation, which is based on a state space model, uses both algorithm partitioning and data partitioning to achieve high efficiency. This implementation emphasizes the reduction of data transfers between the processors and requires all the data transfers to be local. The system accepts the input as a raster scan data stream, which reduces the hardware required for the input data control block. With proper ordering of computations, a matrix transposition between two matrix-by-matrix multiplications, which is required in many DCT systems, can also be removed.<>