摘要
主要介绍了通过基于肤色信息的彩色图像分割的方法,利用预先设定的肤色建模知识进行人脸区域的检测及区域标定。在此基础上,利用给定的人的眼睛、嘴巴的先验色度信息进行眼睛和嘴巴的匹配检测,并最终实现双眼中心的标定。通过对大量图片的实验证明,此种方法速度快,实时性高,算法简单,运算量小,从而达到精度和速度相平衡的人脸以及特征信息检测的目的。
In today's era of rapid development of information technology,the pursuit for the balance of precision and speed detection face detection system has become the key to achieving.In order to realize eyes center calibration ultimately and reach to the Purpose of balance between accuracy and speed of face and feature information detection,this article mainly puts up face region detection and region labeling based on color information of color image segmentation method,using Pre-set color modeling knowledge,Through testing a large number of images show that this method is faster,real-time high,simple algorithm,and computational load,on this basis,executes matched detection for eyes and mouth making use of a given prior chroma of human eyes and mouth.
作者
侯智卿
李仰军
王高
Hou Zhiqing;Li Yang-jun;Wang Gao(Key Laboratory of Instrumentation Science&Dynamic Measurement Ministry ofEducation,North University of China,Jiancaoping District,Taiyuan,Shanxi Province 030051,China)
出处
《电子测试》
2011年第4期41-44,共4页
Electronic Test
关键词
快速人脸检测
肤色模型
定位
精度与速度平衡
fast face detection
skin-based model
locate
Balance between accuracy and speed
