摘要
为了减少大场景点云文件无损压缩时间,避免数据处理过程中人工操作造成的时间浪费和输入失误,本文引入LASzip点云无损压缩算法和OpenMP并行编程技术,提出了在单机多核CPU系统环境下基于LASzip的大场景点云文件多核并行批量压缩算法。实验结果显示,与串行算法相比,并行算法在压缩比不变的情况下压缩时间明显下降,取得了很好的加速比。当线程数为8时,压缩和解压缩加速比分别达到4.27和3.99,优化效果显著。
In order to reduce the lossless compression time of the point cloud data files in large scene,avoid the time consuming and inputting errors caused by manual operation during data processing,this paper introduces LASzip lossless compression algorithm and the technology of OpenMP-based parallel computation,proposes a multi-core parallel batch compression algorithm for point cloud data files in large scene based on LASzip algorithm within the environment of the single machine of multi-core CPU.The experimental results show that compared with original serial algorithm,the parallel algorithm decreases obviously the compression time when the compression ratio is constant,achieving good speedup.The proposed optimization algorithm's compression and decompression time acceleration ratios reach 4.27 and 3.99,respectively,when 8 threads are used,the optimization effects are remarkable.
出处
《太原理工大学学报》
北大核心
2017年第6期973-977,共5页
Journal of Taiyuan University of Technology
基金
国家自然科学基金资助项目(61572345)
国家科技支撑计划项目课题(2015BAH37F01)
