A mass medical image retrieval system based on Hadoop
Fan Xu Sheng
In order to improve the efficiency of mass medical image retrieval, the defect of single node medical image retrieval System This paper presents a mass medical image retrieval system based on Hadoop. Firstly, using Brushlet transform and local two-value mode algorithm to extract medical sample image features and storing image feature inventory in Hadoop Distributed File System (HDFS) And then using map to match the features of the sample image and feature library, reduce receives the calculation results of each map task and sorts by similarity degree. Finally, the optimal retrieval results of medical images are found according to the order results. The experimental results show that the medical image retrieval system of Hadoop reduces the image storage and retrieval time and improves the image retrieval speed compared with other medical image retrieval systems.
A mass medical image retrieval system based on Hadoop