| 一种基于GPU的二维离散多分辨率小波变换加速方法 |
| |
| 引用本文： | 刘磊,张子佳,刘雷,张睿.一种基于GPU的二维离散多分辨率小波变换加速方法[J].吉林大学学报(理学版),2015,53(2):267-272. |
| |
| 作者姓名： | 刘磊 张子佳 刘雷 张睿 |
| |
| 作者单位： | 1. 吉林大学 计算机科学与技术学院, 长春 130012； 2. 中国科学院 计算技术研究所, 北京 100190 |
| |
| 基金项目： | 国家高技术研究发展计划863项目基金(批准号:2012AA010902);国家重点基础研究发展计划973项目基金(批准号:2011CB302500) |
| |
| 摘 要： | 针对传统CPU平台下小波变换算法难满足当前高分辨率、大数据规模下的实时性要求,提出一种基于GPU的并行小波变换算法,并通过改善Local Memory访存数据的局部性和增加Global Memory访存带宽的优化技术,利用多Kernel并行提高多种分辨率下小波变换的性能.实验结果表明,与CPU串并行版本相比,GPU并行优化算法在高分辨率变换情况下,加速比最高可达30~60倍,可满足对变换实时性的要求.
|
| 关 键 词： | 小波变换 多分辨率 GPU加速 |
| 收稿时间： | 2014-05-09 |
A Method of GPU Based Accelerating 2DMulti-resolutions Discrete Wavelet Transform |
| |
| LIU Lei
,
ZHANG Zijia
,
LIU Lei
,
ZHANG Rui.A Method of GPU Based Accelerating 2DMulti-resolutions Discrete Wavelet Transform[J].Journal of Jilin University: Sci Ed,2015,53(2):267-272. |
| |
| Authors: | LIU Lei ZHANG Zijia LIU Lei ZHANG Rui |
| |
| Institution: | 1. College of Computer Science and Technology, Jilin University, Changchun 130012, China;2. Institute of Computing Technology, Chinese Academy of Sciences, Beijing 100190, China |
| |
| Abstract: | Since the classical wavelet transform algorithm on CPU hardly meets the real time performance requirements, especially dealing with large scale data in high resolution, we presented a GPU based parallel wavelet transform algorithm, which improves the locality of local memory access and increases the bandwidth of global memory access. It uses multi kernel to improve the performance in the case of multi resolutions. The experiment results show that compared to the performance of a classical algorithm on CPU, GPU gains the speedup of 30—60, accordingly, it can satisfy the real time requirements for transformation. |
| |
| Keywords: | wavelet transform multi resolutions GPU acclerate |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《吉林大学学报(理学版)》浏览原始摘要信息 |
| 点击此处可从《吉林大学学报(理学版)》下载免费的PDF全文 |
|
