which is based on existing fast DCT algorithm and has fully utilized the parallelism of very long instruction word (VLIW) architecture to achieve high speed performance.Simulation results show that this method can reduce more than 73% and 24% computing cycles
compared with normal and fast DCT computation algorithm respectively.