摘要
分析集群环境下分布式存储编程模型和多核CPU环境下共享存储编程模型各自的优缺点,采用结合集群和多核CPU的并行环境来取长补短;并研究其在等高线提取中的相关并行算法,其中以建立三角网和跟踪等高线作为共享存储并行的研究实例;最后通过实验验证了该并行方案的可行性。
After analyzing the advantages/disadvantages of distributed memory programming model in cluster environment and shared memory programming model in multi-core CPU respectively,a parallel method,combining clustering and multi-core,is proposed to enhance the strengths and eliminate the weaknesses.Then the relevant parallel algorithms of extracting contours are studied,of which establishing TIN-Mesh and tracking contours are as study cases of parallel shared memory.At last,the feasibility of the parallel method is verified by experiments.
出处
《计算机工程与应用》
CSCD
北大核心
2010年第17期5-7,44,共4页
Computer Engineering and Applications
基金
国家高技术研究发展计划(863) No.2006AA12Z101
No.2009AA12Z107
国家重点基础研究发展规划(973) No.2009CB724007
集美大学科研基金 No.4412Z80176~~
关键词
机载激光雷达
等高线
并行
多核CPU
集群
Light Detection And Ranging（LiDAR）
contours
parallel
multi-core CPU
cluster
