| 基于极小值区域的人眼定位新算法 |
| |
| 引用本文： | 刘莎. 基于极小值区域的人眼定位新算法[J]. 微型机与应用, 2011, 30(20): 37-39,43 |
| |
| 作者姓名： | 刘莎 |
| |
| 作者单位： | 仰恩大学计算机与信息学院计算机系,福建泉州,362014 |
| |
| 摘 要： | 提出了一种基于灰度图像极小值区域的快速鲁棒的人眼自动定位方法。首先,利用极小值区域进行眼睛粗定位,得到一组候选眼睛位置;再利用眼睛对的几个自然约束条件进行粗筛选;最后利用PCA方法验证眼睛对,得到唯一的候选眼睛对。对BioID图像集进行测试,结果表明,利用该算法进行眼睛定位的成功率和定位精度较高。
|
| 关 键 词： | 眼睛定位 极小值区域 主成分分析 |
Eye location based on minimum extremal regions |
| |
| Liu Sha. Eye location based on minimum extremal regions[J]. Microcomputer & its Applications, 2011, 30(20): 37-39,43 |
| |
| Authors: | Liu Sha |
| |
| Affiliation: | Liu Sha (Department of Computer,College of Computer and Information,Yang-En University,Quanzhou 362014,China) |
| |
| Abstract: | A novel fast eye-locating method for the gray image is proposed. This algorithm involves three main steps. Firstly, MERs （Minimum Extremal Regions） is extracted to roughly locate eye candidates. Then, nature restraints are applied to filter out false eyes. Finally, PCA method is used to find out the right eye-pair. Experiments on BiolD image databases prove that the pro- posed eye-locating algorithm has hlgher success rate and positioning alluraly. |
| |
| Keywords: | eye location minimum extremal region PCA |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
