Face Detection of Color Image Based on Skin Color and AdaBoost Algorithm

Song Yiwei,Wang Xiu,Zhao Xuezhu,Zhu Xuefeng
DOI: https://doi.org/10.3969/j.issn.1674-2605.2009.01.003
2009-01-01
Abstract:This paper proposes a novel face detection method combined skin color detection and AdaBoost algorithm. First, skin color segmentation is used to detect skin regions and candidate face regions are formed in terms of the statistical characteristics of human face. Then these candidate face regions are scanned by cascade classifier which is consisted of many strong classifiers based on AdaBoost algorithm for more accurate face detection. The experimental results show that the new approach can detect face with high speed and low false acceptance. It can be effectively applied to the cases of multi-pose and multi-face images.
What problem does this paper attempt to address?