云平台上基于图像特征索引的并行检索系统技术研究

为了提高建立索引、检索图像的速度,提出云架构上基于图像特征索引的并行检索系统.该检索系统主要有3个模块:海量小图片分布式存储(Store)、并行建立图像特征索引(Indexing)、并行图像检索(Retrieve).在Store模块中提出针对海量图片的合并存储,Indexing模块中提出索引缓存模式,避免重写索引的输出接口,Retrieve模块中对索引进行分片管理,以及并行检索.实验结果表明,相对于其他图像检索系统,基于图像特征索引的检索系统有效减少了图像特征索引建立时间,缩短了图像的检索时间,提高了图像检索速率.

PARALLEL RETRIEVAL SYSTEM BASED ON IMAGE FEATURE INDEX

In order to improve the speed of indexing and retrieving images, in this paper,we propose the parallel retrieval system based on image features index in cloud.The retrieval system mainly consists of three parts:the distributed store of massive small images(Store),building index of image features in parallel(Indexing),image retrieval in parallel(Retrieve). In the part of Store,it proposes the combined storage for massive small images;in the part of Indexing,in order avoid rewriting the output interface of index,it proposes the mode for caching index;in the part of Retrieve,by the fragmented management for features index ,it propose the image retrieval in parallel.The experimental results show that the retrieval system reduces the time for indexing and retrieving effectively,it improves the speed for image retrieval.