A personal identity annotation overlay system using a wearable computer for augmented reality

  • Authors:
  • Jong Bae Kim

  • Affiliations:
  • Sch. of Electr. & Comput. Eng., Georgia Inst. of Technol., Atlanta, GA, USA

  • Venue:
  • IEEE Transactions on Consumer Electronics
  • Year:
  • 2003

Quantified Score

Hi-index 0.43

Visualization

Abstract

There is surely no doubt that computers play an important role in our life. The use of computers ranges from convenient tools like calculators to various lives saving equipment. They help to save a great amount of human effort in many areas. To accomplish special tasks in different workplaces, it would be of great advantage to develop proper devices and to use wearable computers to meet the goals successfully. That is why we use the support of different kinds of computers such as the Global Positioning System (GPS), personal digital assistant (PDA), digital bar code reader, etc. Using a portable or wearable computer, we can work outside our offices or even in bed. Thus, our work environment has changed considerably. In many new environments, computers can help us to increase our perception and interaction abilities. Augmented reality (AR) is a good example of a tool used to enhance a human's perception and interaction abilities. AR provides a natural method for presenting computer-generated information by overlaying graphics into a view of the real world. This paper describes a personal identity annotation system using a wearable computer for an AR environment, which allows its user to recognize unknown persons. For implementing, the proposed system uses skin-color pixels coupled with shape and face features to detect and track faces in images taken by a wearable camera. Faces in a view are detected by searching for connected regions of skin color pixels. For each region big enough, an elliptic area is determined with the highest image gradient magnitude around the perimeter and the highest number of enclosed skin color pixels. The face regions enclosed by the determined ellipses are then used by the system to recognize multiple faces based on the face similarity. Experimental results show that the proposed system has both quality and performance to be used as a guiding system for visitors in various equipments.