Secure and Privacy-assured Image Service Outsourcing in Cloud Computing

With the advancement of information and computing technology, large-scale image datasets are being exponentially generated today. Along with such data explosion is the fast-growing challenge to efficiently and effectively store, maintain, and share them from data owners to a large number of data users. Cloud, with its robust yet economical computing resources, provides a promising platform for hosting such image management systems. However, for the image service outsourcing paradigm to become truly successful, there are still fundamental and critical challenges yet to overcome. Firstly, because the cloud is a public environment operated by external third-parties usually outside the data owner/users’ trusted domain, the outsourcing design has to be privacy-protecting and sometimes mandatorily provide legal compliance to various privacy regulations. Secondly, due to the high-dimensionality and large-scale of the image datasets, it is both necessary and desirable for the image service design to be as efficient and less resource-consuming as possible, in order to keep the cloud economically attractive.To address these fundamental challenges, we propose to investigate and prototype a novel secure image service outsourcing architecture that aims to exploit techniques from different domains and take security, complexity, and efficiency into consideration from the very beginning of the service flow. Particularly, we propose to research our service design under the compressive sensing framework, a recent data sensing/sampling paradigm known for its simplicity of unifying the traditional sampling and compression for image acquisition. Data owners only need to outsource compressed image samples to cloud for reduced storage overhead and simplified local sensing. We will explore a privacy-preserving image recovery service, where data owners/users can securely harness the cloud as a central hub responsible for image recovery from sample processing on demand, without revealing information on either the image samples or the recovered image content. With more and more image samples centralized to cloud, we further investigate a privacy-preserving content based image retrieval service, which allows users to selectively retrieve encrypted image samples from cloud without revealing their private interests. The proposed service designs will naturally interoperate and significantly boost the wide application spectrum of “secure computing outsourcing”. The research result can be easily adopted by a diverse set of industrial applications like MRI in health care system, remote sensing in geographical system, and even military image sensing in mission critical contexts. The project will also contribute to the educational plan, including curriculum development, undergraduate/graduate student training and research mentoring.