三维小波变换在IBR数据压缩中的应用  

Application of 3D Wavelet Transform in IBR Data Compression

作　　者：章勤[1] 聂江武[1] 郑然[1] 王海岛[1] 

机构地区：[1]华中科技大学集群与网格计算实验室,湖北武汉430074

出　　处：《计算机工程与科学》2005年第9期37-39,53,共4页Computer Engineering & Science

基　　金：中国教育科研网格计划资助项目(CG2003GA001)

摘　　要：针对基于图像的绘制IBR场景数据的特点,本文给出了一种有效的图像序列压缩方法,以降低这种场景数据传输量。本文首先对图像序列进行“滑动窗口”的选帧,然后对图像序列进行基于整数提升的三维小波变换,最后利用嵌入式零树小波编码对小波系数进行量化编码,从而实现对IBR场景数据的压缩。实验结果显示,这种方法达到了良好的压缩效果。IBR is a new scheme which can represent scenes, and it offers a quick way to model and render a realistic 3D scene. But the data produced by this method is huge. In this paper, we give a scheme to compress the image array. Our scheme first selects the image array we need urgently, and then applies a lifting-based 3D wavelet transform on the selected image array, and finally quantizes different bands with different precisions and entropy codes for each band. The experiments show that this scheme can achieve very good results.

关 键 词：IBR 三维小波变换 提升格式 嵌入式零树小波编码 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]