自适应肤色检测算法的研究

Study on adaptive skin color detection algorithm

针对Hsu提出的椭圆肤色检测模型的局限性,提出了一种自适应肤色检测模型。在Hsu的统计模型基础上,根据光照条件自适应改变其椭圆模型的长短轴,以取得最好的肤色检测效果。实验结果表明,本文算法有较强的光照适应性,所提出的自适应肤色检测模型效果优于Hsu的椭圆模型。

In this paper, an adaptive skin color detection model is propose to improve complexion detection ability of the elliptic model proposed by Hsu et al. Better complexion detection effect is obtained by adapting the sizes of the major and minor ax- es according to the illumination. Experimental results have demonstrated that the proposed model can adapt the changing of illuminating condition and its complexion detection ability is better than the elliptic model suggested by Hsu et al.