期刊名称：International Journal of Security and Its Applications
印刷版ISSN：1738-9976
出版年度：2015
卷号：9
期号：8
页码：1-14
DOI：10.14257/ijsia.2015.9.8.01
出版社：SERSC
摘要：With the growing popularity of cloud computing, more and more data owners are willing to outsource their data to the cloud. However, private data should be encrypted before outsourcing for security requirements, which obsoletes data utilization like content-based image retrieval. In this paper, we propose a secure similarity image search scheme, which allows data owners to outsource their encrypted image database to the cloud server without revealing the real content of images. The proposed scheme supports both global and local feature based image retrieval under various distance metrics, such as earth mover's distance. Firstly, the data owner extracts either global features or local features from images to represent the images. Then, these features are used to generate a searchable index. Finally, both image database and searchable index are encrypted before outsourcing to the cloud server. When a query image coming, the data user extracts feature from the query image and generates the search trapdoor. The trapdoor is then sent to the cloud server and used to compare the similarity with the searchable index. Extensive experiments are conducted to show the efficiency and applicability of our proposed similarity image search system.
