Hierarchical image-scrambling method with scramble-level controllability for privacy protection

Abstract

Privacy-protecting technology is essential in this surveillance society. The applications of cameras widely vary, e.g., crime surveillance, monitoring of an environment, and marketing. Furthermore, the scale of surveillance systems is predicted to become more diverse (from home area networks to wide area networks) due to the decrease in size and price of cameras. Therefore, a simple privacy protection system that does not require central servers or large databases is needed. Scrambling private information in a captured image can be a solution to simplifying a system. We propose an image-scrambling method for bitmap and JPEG formatted images to private information. Our method enables access control by providing keys to authorized individuals. They cannot view private information that they do not have permission to access. The image's format is retained; therefore, no special viewer is necessary in display-only console. Experimental results suggest that scramble level can be controlled linearly by using parameters (three for JPEG formatted image, and one for bitmap image). We also developed a demo system for this method and confirm that this method can be applied to embedded systems such as those equipped with surveillance cameras.