Privacy Enhanced Pixel-Level Image Processing in the Clouds

Image processing and storage are enormously resource intensive tasks that can benefit from cloud computing. Lack of robust mechanisms for controlling the privacy of the data outsourced to clouds is one of the concerns in using clouds for image processing. This paper presents a new image encoding scheme that enhances the privacy of the images outsourced to the clouds while allowing the clouds to perform certain forms of computations on the images. Our prototype shows the feasibility of performing a class of image processing tasks on images encoded for privacy.