用黑斑的椭圆性质在二值化图像中探测眼睛  

Using ellipse properties to find eyes in binary image

作　　者：范宏深[1] 倪国强[1] 申会堂[2] 

机构地区：[1]北京理工大学光电工程系,北京100081 [2]中国船舶重工集团公司第718研究所,河南邯郸056027

出　　处：《光学技术》2001年第4期307-309,共3页Optical Technique

摘　　要：在人脸探测过程中 ,双眼的探测非常重要 ,这是因为人的眼睛是人脸上最容易探测到的特征部位。通过双眼探测还可以对人脸进行定位 ,使人脸的尺寸归一化。探讨了一种基于二值化图像中黑斑几何形状分析的双眼探测方法。因为人眼的形状比较接近椭圆 ,故在研究黑斑的几何性质时 ,重点研究黑斑的椭圆性质。所谓黑斑的椭圆性质是指按照黑斑的面积和二阶矩计算出的椭圆参数。椭圆的主要参数有椭圆的面积、质心、长短轴的长度和方向角 (椭圆长轴与水平方向的夹角 )。为了描述黑斑与椭圆的近似程度 ,定义了椭圆硬度的概念。根据眼睛黑斑椭圆参数的测试数据建立了探测眼睛的数学模型 。Eyes detecting is very important in face detection Eyes are the easy detecting part in a face and once the eyes are detected the face can be located and normalized In this paper a new method to detect eyes is discussed which is based on the geometric analysis of all the objects in a binary image using ellipse properties including its area, its center position, its orientation, its major axis length and minor axis length All the objects in a binary image have a corresponding ellipse that is calculated by their second moments and areas In order to describe how an object matches an ellipse, a new phrase, ellipse solidity, has been coined here By measuring the experiment dada of all the ellipse properties, a model is built and some eyes detecting results are given

关 键 词：人脸探测 人眼探测 椭圆性质 黑斑 二值化图像 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置] TP391.41[自动化与计算机技术—控制科学与工程]