摘要
分析了MATLAB并行计算工具箱中各部件的关系,在Windows环境下搭建了并行计算集群;采用滑动邻域操作对声纳图像进行了对比度增强处理;重点介绍了MATLAB下的数据并行编程,利用分布式数组设计了集群环境下的图像增强并行算法。实验结果表明,MATLAB强大的内部函数使得并行计算易于实现,有效地提高了图像处理的实时性。
The relationship between different components in MATLAB parallel computing toolbox is analyzed and parallel com- puting cluster is established in the environment of Windows. To enhance the contrast of sonar images, moving neighborhood opera- tion is adopted. This paper especially focused on data-parallel programming and it employed distributed arrays to design parallel al- gorithm of image enhancement under cluster environment. The result demonstrates that strong internal funetion in MATLAB make parallel image algorithm easy to implement and improve the real-time of image processing effectively.
出处
《微型机与应用》
2013年第8期32-34,共3页
Microcomputer & Its Applications
基金
国家自然科学基金资助项目(41176076)
关键词
对比度增强
并行计算工具箱
滑动邻域
声纳图像
数据并行编程
分布式数组
contrast enhancement
parallel computing toolbox
moving neighborhood
sonar images
data-parallel programming
distributed array
