CPU/GPU异构系统下高光谱遥感影像降维多级协同并行计算方法及优化策略

项目名称： CPU/GPU异构系统下高光谱遥感影像降维多级协同并行计算方法及优化策略

项目编号： No.61272146

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 自动化技术、计算机技术

项目作者： 周海芳

作者单位： 中国人民解放军国防科学技术大学

项目金额： 74万元

中文摘要： 如何对高光谱信息进行高效处理是遥感领域近年来的研究热点，高光谱影像降维是做好后续地物分析识别的一个关键步骤。由于降维计算是一个典型的计算密集和访存密集型过程，计算复杂度很高，采用传统的串行处理模式，已无法满足军事、农林等高端应用的实时性需求。通用CPU和专用GPU相互配合的异构系统可以满足应用对计算资源的不同需求，是未来高性能计算机体系结构最有前景的发展方向之一。本项目将针对高光谱遥感影像降维计算方法，以线性降维方法为牵引，重点研究基于流形学习的非线性高光谱影像降维算法，建立性能分析模型，研究提出基于新型CPU/GPU异构体系结构的多级协同并行算法，归纳此类应用面向该体系结构特征的一般优化方法，并反向指导降维方法在可并行度和可扩展性方面的创新，突破原有算法的"加速比墙"，出原创成果，有效提高高光谱遥感影像处理的业务水平，应用前景广阔。

中文关键词： GPU；异构系统；高光谱影像；降维；并行算法

英文摘要： How to efficiently process hyperspectral information has become the research focus of remote sensing area. Dimensionality reduction of hyperspectral image is the key step of the following analysis and identification of landscape. Dimensionality reduction is a typical computing intensive and memory access intensive process with high complexity. Realizing data reduction in serial mode is impossible to satisfy the real-time need of many applications, such as military and agriculture. Heterogeneous system with general CPU and special GPU can satisfy different needs for computing resources, and is one of the most promising developments of future high performance computer architecture. This project, focused on dimensionality reduction of hyperspectral image, will start from linear reduction algorithms and primarily study the nonlinear reduction algorithms of hyperspectral image based on manifold learning method. A performance analysis model will be built and a series of multilevel cooperative parallel algorithm for novel CPU/GPU heterogeneous architecture will be proposed. Then, the project summarizes the general optimization method of this kind of applications on the CPU/GPU heterogeneous architecture, which will provide some innovative ideas of improving parallelism and scalability of dimensionality reduction algor

英文关键词： GPU；heterogeneous system；hyperspectral image；reduction；parallel algorithm