基于三维小波的大规模体数据压缩方案

Wavelet-Based3D Compressing Scheme for Large Volume Data

可视化大规模体数据在科学和工程领域一直被认为是困难的。特别是对于那些经常要求非常大的运行时间存储空间的数据尤其如此。文章讨论了基于小波理论的对于大规模体数据的有效的三维压缩方案。在设计该压缩方时,对两个重要参数进行了折衷:高的压缩率和快速运行时间随机访问。可视化人体数据集的实验结果表明此方法取了相当好的压缩率,另外,由像素值的运行时间重建引起的开销达到了最小值。这种三维压缩方案在开发用于大规模数据的交互式可视化系统时非常有用,并且使得更多的用户,象基于个人电脑或具有有限容量的低端工作站都可能运可视化技术。

Visualizing very large volume data has been recognized as a task requiring great effort in a variety of sci-ence and engineering fields.In particular,such data usually places considerable demands on run-time memory space.This paper describes an effective3D compression scheme for very large volume data that exploits the power of wavelet theo-ry.In designing this compression method,it has compromised between two important factor:high compression ratio and fast run-time random access.This experimental results on the Visual Human data sets show that the method achieves fairly good compression ratios.In addition,it minimizes the overhead caused during run-time reconstruction of voxel val-ues.This3D compression scheme will be useful in developing many interactive visualization systems for huge volume data,and will make visualization technology accessible to a much wider range of users,as it can be based on personal computers or low-end workstations with limited memory.