基于局部对称性及灰度统计特征的人眼定位方法

Eye locating algorithm based on local symmetry and statistic feature of gray value

为了兼顾人眼定位方法的准确率和实时性,提出基于局部对称性和灰度统计特征的人眼定位方法。该方法结合了模板匹配与对称变换的思想,利用了人眼灰度统计信息,取得了较好的效果。通过仿真实验表明:算法在满足算法定位准确率的同时,具有较好的实时性,因为其算法复杂度相对于模板匹配法和基于学习的方法大大降低。

To balance between accuracy and real time of eye locating algorithm,an eye locating algorithm is proposed in this paper based on local symmetry and statistic feature of gray value.The algorithm combines the advantages of the template matching algorithm and the symmetry transformation algorithm,and uses the statistic feature of gray value of eyes.The case study shows the proposed algorithm achieves a better real time performance while the accuracy is guaranteed.This is because the proposed algorithm is less complicated than that of other algorithms.