一种基于YCgCr色彩空间的人脸检测方法

摘要

Face detection is one of the important research directions in Computer Vision.An face detection algorithm based on the Gaussian model and new color space YCgCr is proposed.Firstly,we compare the two color spaces of YCgCr and YCbCr,and find the clustering characteristics of skin color in the the YCgCr is better than YCbCr color space.Then get the interchange area through binary image.Finally,find the face according to one aspect ratio to in addition to interference region,obtain accurate facial images.Experimental results show that the algorithm is effective.and fast%人脸检测是计算机视觉研究的重要研究方向之一,提出了一种基于新色彩空间YCgCr的人脸检测算法。首先比较了YCgCr间和YCbCr色彩空间,得出YCgCr色彩空间的在人脸处理方面更有优势,然后通过二维高斯分布提取肤色区域,对二值化的图像应用区域增长算法求取最大联通区域,得到准确的面部图像,实验结果表明该算法是快速而有效的。