A Secure Image Outsourcing Using Privacy-Preserved Local Color Layout Descriptor in Cloud Environment
J. Anju, R. Shreelekshmi
Abstract
Secure storage and content-based retrieval of images on third-party servers without compromising privacy and efficiency is a challenging research problem. In this paper, we propose novel encryption and indexing schemes for secure storage of sensitive images on the cloud server and secure retrieval of the images by authorized image users. The proposed image encryption preserves the Local Color Layout Descriptor for facilitating secure outsourcing of privacy-preserved feature extraction and indexing to the cloud server. The proposed indexing represents each image using a dual image signature generated from encrypted Local Color Layout Descriptors extracted from encrypted images. The scheme employs a novel relevance score incorporating multiple distance metrics on dual image signatures and image matrix rank for better retrieval precision and speed. Experimental evaluation shows that the proposed scheme achieves a mean average precision of 0.68555 with a search time of 0.026 seconds outperforming all the state-of-the-art schemes for image outsourcing. Security analysis is also performed which reveals that the scheme is more secure than the state-of-the-art schemes.