文章基本信息

标题：Stream Processing To Solve Image Search by Similarity
本地全文：下载
作者：Jair Lobos ; Verónica Gil-Costa ; Nora Reyes 等
期刊名称：Journal of Computer Science and Technology
印刷版ISSN：1666-6046
电子版ISSN：1666-6038
出版年度：2015
卷号：15
期号：2
出版社：Iberoamerican Science & Technology Education Consortium
摘要：The classic use of Stream Processing platforms enablesworking with data in real time, which allows you togenerate data analysis quickly attending to a decisionmakingprocess. However, you can use these platformsfor other applications such as indexing and subsequentuse of similarity search objects in a database. The imagescan be displayed on a metric space, which has featuresthat allow rules to discard a not similar image quicklywithout making costly computations. This paper presentsthe use of a Stream Processing platform to index imagesgenerated by different users. For this, it is necessary torepresent these images by vectors containing differentMPGE-7 features. This paper shows a Stream Processingplatform using its processing elements (PEs) in parallel tospeed up the operations involved in the indexconstruction.
关键词：Stream Processing; Metrics Spaces; MPEG-7;Sparse Spatial Selection