一种由粗略到精确的瞳孔定位算法

摘要

Aiming at pupil precise localization algorithm is too complicated in present stage, put forward a pupil precise localization algorithm from rough to precise. First, according to grayscale integral projection of the face,the general location of the eyes can be determined quickly. Then use the Hough transform to locate the exact position of eye pupil center. Next, use Harris corner detection algorithm find the corner of the eye. Finally, determine the exact position of the pupil through the coordinate location of the pupil and eye corner. The proposed algorithm tested with the MATLAB, it is proved to be a simple, accurate positioning algorithm.%针对现阶段实现瞳孔精确定位算法过于繁琐的问题，提出一种由粗略到精确的瞳孔精确定位算法。首先用灰度积分算法粗略找出人眼部位在人脸上的大概位置，然后用Hough变换定位出人眼瞳孔中心的精确位置，接着用Harris角点检测算法对眼角点进行定位，最后通过瞳孔与眼角点的相互坐标定位出人眼瞳孔的精确位置。通过matlab对所提算法进行检验，证明了该算法是一种实现简单、定位精确的算法。