人眼的识别是计算机人脸识别和智能监控中的重要部分。本文提出了一种基于灰度投影

曲线和边缘检测进行人眼定位的新算法。首先利用原图的垂直灰度投影曲线确定人脸的左右边界, 之

后利用人脸区域的水平灰度投影确定头顶至鼻中部的上下边界, 然后利用人脸比例模型估计人眼的大

概位置, 分析人眼区域的边缘图象, 给出人眼的确切位置。实现表明, 该算法简单、快速, 鲁棒性强-Identification of the human eye is the computer face recognition and intelligent control of the important part. This paper presents a projection based on gray curve and edge detection to locate the new algorithm for the human eye. Firstly, original gray projection curve to determine the vertical face of the left and right boundaries, and then use the face region to determine the level of gray projection head to the nose and down the middle of the border, then use the model to estimate the proportion of people face around the eye position analysis the edge of the eye region image, given the exact location of the human eye. Realize that the algorithm is simple, fast and robust