摘要
如何高效地进行视频数据压缩是一直备受关注的研究问题。张量是高维数据的自然表示,张量紧凑表可以大幅降低原始数据维数,且能非常近似地恢复原数据。文中根据张量紧凑表示概念提出张量迭代Tucker-ALS算法,并将该算法应用至视频压缩中,取得较好的压缩效果。通过测试序列仿真并运用BD-rate比较方法进行压缩性能评估,相比于目前成熟的H.264算法,文中所提出的迭代Tucker-ALS算法在低码率时性能有所改善,对于纹理类视频性能改善显著。
Efficient video data compression is always a research concern. Tensor is a natural representation of high dimensional data, and tensor compact representation can greatly reduce the dimension of original data and recover the original data very well. Base on the concept of compact representation of tensor, we propose the tensor iterative Tucker - ALS algorithm for better video compression. The comparison of the simulation results with those by the BD - rate method shows that the proposed tensor iterative Tucker - ALS algorithm can improve the performance compared with the mature H. 264 algorithm at low bit rates with significantly improved performance for texture video.
出处
《电子科技》
2017年第5期1-4,共4页
Electronic Science and Technology
基金
国家自然科学基金(61671361)
