基于MapReduce数字图像处理研究

随着海量图像数据的增加，使得需要处理的数据规模越来越大，为了解决在处理海量数据信息时所面临的存取容量和处理速度的问题，在深入研究MapReduce大规模数据集分布式计算模型的基础之上，本文设计了基于MapReduce实现对数字图像并行化处理。实验结果表明：运行在Hadoop集群上的基于MapReduce并行化算法具有数据节点规模易扩展、处理速度快、安全性高、容易实现等特点，能够较好地满足海量数据图像的处理的要求。

Research of digital image processing based on MapReduce

With the increase of mass image data, makes the need to deal with the data size is bigger and bigger, in order to solve the facing access when dealing with huge amounts of data information capacity and processing speed, the further study of graphs large-scale distributed computing model based on the data set, in this paper, based on graphs design for digital image parallel processing. Experimental results show that run on Hadoop cluster graphs based on parallel algorithm is data node size easy extension, fast processing speed, high security, easy to implement, can well meet the requirements of mass data processing of the image.