| 基于GPGPU的数字图像并行化预处理 |
| |
| 引用本文： | 宋晓丽,王庆.基于GPGPU的数字图像并行化预处理[J].计算机测量与控制,2009,17(6):1169-1171. |
| |
| 作者姓名： | 宋晓丽 王庆 |
| |
| 作者单位： | 西北工业大学,计算机学院,陕西,西安,710072 |
| |
| 基金项目： | 国防预研项目，国家高技术研究发展计划(863计划) |
| |
| 摘 要： | 首先简要介绍了统一设备架构CUDA(Compute Unified Device Architecture)技术的背景、特点、内存模型,利用通用计算图形处理单元GPGPU(General Purpose GPU)及CUDA技术,实现了图像直方图均衡化和薄云去除的并行化处理,与传统的基于CPU的方法相比,两个基于GPGPU的图像预处理操作的执行效率分别提高了40倍与80倍左右,在大规模实时性图像处理操作中,有很大的实用价值。
|
| 关 键 词： | GPGPU CUDA 直方图均衡 薄云去除 并行计算 |
Digital Image Parallel Pre-Processing Based on GPGPU |
| |
| Song Xiaoli,Wang Qing.Digital Image Parallel Pre-Processing Based on GPGPU[J].Computer Measurement & Control,2009,17(6):1169-1171. |
| |
| Authors: | Song Xiaoli Wang Qing |
| |
| Affiliation: | School of Computer Science;Northwestern Polytechnical University;Xi'an 710072;China |
| |
| Abstract: | Parallel processing of histogram equalization and removal of light cloud for gray scale images are implemented by GPGPU and CUDA technique.The efficiencies of two image pre-processing operations simulated on GPGPU can be improved up to 40 and 80 times compared to the sequential methods based on CPU,respectively.This approach can be redounded to real-time applications for large scale image processing and analysis. |
| |
| Keywords: | GPGPU CUDA |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
