期刊论文

Eye State Detection Based on LBP Template Matching

中文

现代计算机：中旬刊

1007-1423

2016

21

35-38

10.3969/j.issn.1007-1423.2016.21.008

湖南省哲学社会科学基金（No.14YBA335）；

本校为第一机构

提出一种基于LBP算子模板匹配的人眼状态检测方法。首先,对实验中用到的图像进行LBP算子计算,对样本图像和原始图像的各个像素点进行相似度计算,得相似度矩阵。之后再通过模板匹配,得出眼睛的精确区域。最后根据一定的比例系数和相应的公式来识别眼睛的状态。实验表明,基于LBP算子的模板匹配可以更好突出眼睛的纹理特征,从而得到更高的正确率。

Presents the state of the eye detection method based on a LBP operator template matching. First, the image used in the experiment LBP operator is calculated for each pixel of the original image and the sample image similarity calculation to obtain similarity matrix. After through template matching, draws precise regions of the eye. Finally, according to a certain scale factor and the corresponding formula to identify the state of the eye. Experimental results show that the template matching based on LBP operator can better highlight...