摘要
湍流退化图像的恢复面临着计算量大、耗时长等问题,为提高算法的有效性、适应性和快速性,主要对红外探测湍流退化图像复原算法的并行处理进行了研究,提出了基于机群计算的并行处理方法。采用并行编程技术,建立基于微机多节点计算的并行机群系统,充分利用各节点的计算资源,提高并行速度,优化并行效率。实验结果表明,所提出的基于机群计算的并行方法十分有效,提高了湍流退化图像复原处理的计算效率。
In order to improve the efficiency, adaptability and speed of the restoration algorithm for turbulence-degraded image with high computation complexity and long consuming time, the parallel processing method of the restoration algorithm is studied in this paper. And a parallel implement method based on cluster computation of the restoration algorithm is proposed. Parallel programming technology is introduced to solve the computation complexity of restoration algorithm, and a parallel system based on cluster computation with many nodes is built. This parallel system can take advantage of the computation resource, improve the parallel speed and optimize the parallel efficiency. Experiment results show that the proposed parallel method based on cluster computation is very effective, which heightens the computation efficiency.
出处
《红外技术》
CSCD
北大核心
2009年第1期57-60,共4页
Infrared Technology
基金
国家自然科学基金资助项目(编号:60572040)
中国博士后科学基金资助(编号:2005038538)
关键词
湍流退化图像
图像复原
并行处理
机群
turbulence-degraded image
image restoration
parallel processing
cluster
