A novel cost-effective skin-tone detector for consumer devices

Skin-tone detection has received considerable attention in recent years and applied in wide range of image processing and computer vision applications. The objective of this research project is to develop a real-time skin-tone detection solution optimized for consumer devices, allows for a deliverance of high detection performance at minimal computational costs. The proposed solution will deliver reliable and robust performance even under varying illumination conditions and unconstrained complex background. Key challenge faced by researchers dealing with skin tone detection includes selection of a model of skin color distribution and a color space. Addressing this challenge requires collaborative research which involves academia and industry. We will carry out comparative evaluation of various skin-tone detectors in terms of detection performance and computational complexity to identify the best algorithms suit for our requirements. Consequently, the proposed research will allow us to develop a low complexity skin-tone detection module, enabling robust real-time detection in consumer grade devices.

Faculty Supervisor:

Dr. Konstantinos N. Plataniotis

Student:

Dohyoung Lee

Partner:

Qualcomm Canada Inc.

Discipline:

Engineering - computer / electrical

Sector:

Information and communications technologies

University:

University of Toronto