Visual Bubble: Protecting Privacy in Wearable Cameras

Abstract

Wearable cameras are being used more frequently in many different consumer applications, including entertainment, law enforcement, and health care. To protect the privacy of the environment and bystanders, we introduce a new visual privacy paradigm called the visual bubble. In contrast to most existing visual privacy schemes, the visual bubble is based on depth estimation to determine the extent of privacy protection. To demonstrate this concept, we built a wearable privacy stereo-camera system using the Raspberry Pi platform. The effectiveness of our system in protecting privacy was demonstrated with experimental results on multiple data sets.