基于MapReduce的高分辨率图像高动态重建技术研究
首发时间:2016-05-13
摘要:本文针对大规模像感器阵列分视场采集系统的高分辨率高动态成像问题,采用高分辨率并行计算的新思路,建立了高效的基于MapReduce的分布式高分辨率图像计算重建与显示系统。本算法有效地克服了现有几种主流高分辨率计算成像方法无法同时兼顾高分辨率和高动态多尺度成像的缺点,通过部署分布式计算集群,使得算法具有很高的时效性。实验表明,本文所提出的方法能够快速准确地实现此类分视场成像系统的高动态高分辨率多尺度成像,具有良好的鲁棒性。
关键词: 图像处理 高分辨率 图像拼接 并行计算 MapReduce
For information in English, please click here
The high dynamic reconstruction of high resolution image based on MapReduce
Abstract:By novely using parallel computing technology in high resolution image computational reconstruction, we established a highly efficient distributed high resolution image reconstruction and display system based on MapReduce. This method can be used in high resolution and high dynamic image reconstruction of the Multi-scale Computational Photography instrument(MMCP) based on large-scale sensor-array.Our algorithm effectively overcomes the weakness of most existing high resolution image computational reconstruction methods which cannot be both high resolution,high dynamic and multi-scale. Our system is of highly efficiency through the deployment of distributed computing cluster. We proved that our system presented in this article can be quickly and accurately to achieve high dynamic and high resolution imaging with great robustness.
Keywords: Image process High resolution Image Stitch Parallel computing MapReduce
论文图表:
引用
No.4686339114511814****
同行评议
共计0人参与
勘误表
基于MapReduce的高分辨率图像高动态重建技术研究
评论
全部评论