High Performance Video Compression Based on Improved VSPIHT Algorithm

摘要

This paper proposes an effective I3D-DWT-VSPIHT video coding method, which reasonably organizes wavelet coefficients by integrated 2D-SPIHT with 3D-VSPIHT, and the video sequence is partitioned into two alternate parts: original frames and residual frames, so the similar coefficients can be compressed to one or fewer symbols by sufficiently using correlation between trees, and the reconstructive image quality as well as memory saving can be improved. On the other hand, by effectively reducing redundant scan time it is available in real-time system. Simulated results show that the quality (PSNR) of reconstructed video at the same compression ratio is improved about 1 ～ 5dB for gray video as compared with existent 3D-DWT-SPIHT. By using the high performance video compression method, available for gray and color video sequences, very low bit rate and high quality of reconstructed image can be obtained.