A Hybrid Cascade Approach for Human Skin Segmentation

Samy Bakheet *

Department of Mathematics & Computer Science, Faculty of Science, Sohag University, P.O.Box 82524 Sohag, Egypt and Institute for Information Technology and Communications, Otto-von-Guericke-University Magdeburg, P.O.Box 4120, 39016 Magdeburg, Germany.

Ayoub Al-Hamadi

Institute for Information Technology and Communications, Otto-von-Guericke-University Magdeburg, P.O.Box 4120, 39016 Magdeburg, Germany.

*Author to whom correspondence should be addressed.


Abstract

Human skin segmentation is fundamental to a wide range of computer vision applications ranging from face recognition, facial expression recognition and gesture analysis to various human computer interaction domains. In this paper, we propose a multistage skin segmentation method built as a cascade of a nonparametric generic model and an adaBoost classifier. Several entities are used to train the adaBoost classifier. Feature vectors fed into the ada Boost contain color information from two different color spaces. Extensive experiments are conducted on two datasets in order to evaluate the performance of the approach. The various results obtained show that the proposed method is a promising approach and it successfully achieves high quality segmentation, while concurrently retaining reasonably low false alarm rates. The comparison of the proposed method with related state-of-the-art competitors reveals the superiority and effectiveness of the proposed method, while maintaining real-time performance.

Keywords: Skin segmentation, adaBoost, gesture detection, pattern recognition.


How to Cite

Bakheet, Samy, and Ayoub Al-Hamadi. 2016. “A Hybrid Cascade Approach for Human Skin Segmentation”. Journal of Advances in Mathematics and Computer Science 17 (6):1-14. https://doi.org/10.9734/BJMCS/2016/26412.

Downloads

Download data is not yet available.