On Extracting Facial Expressions for the Virtual Image Masking

Abstract

Since PC camera based net-meeting and video chatting program have been introduced, users become to enjoy real-time communication with each other through PC as having face to face meeting. This paper proposes the virtual image masking for the net-meeting users on internet. Net-meeting users adapted by the virtual image masking can hide their face on the other party’s monitor and show only their facial expressions instead of face, whose shapes could be a tiger or a spider-man. For implementation of the virtual image masking, we first detect a facial area to adapt skin color method with the standard of Asian’s skin color. And we find positions of eyes on the facial area, then, extract the user’s expression like winking or blinking. This virtual image masking would be useful to protect user’s privacy on the open communication environment like internet.