| 彩色图像中的人眼对定位 |
| |
| 引用本文： | 顾晓波,Wang Yaoming.彩色图像中的人眼对定位[J].计算机应用与软件,2008,25(8). |
| |
| 作者姓名： | 顾晓波 Wang Yaoming |
| |
| 作者单位： | 上海师范大学数理信息学院,上海,200234 |
| |
| 摘 要： | 提出了一种彩色图像中人眼对定位的方法.为了减少SVM验证眼睛的搜索范围,首先,采用了滞后阈值法对肤色相似度图分割的方法,进一步,利用眼睛的色度、亮度、梯度等信息以及区域规则确定候选范围,接着根据眼睛的几何分布排除SVM误检的区域,最后用投影函数大致定位出眼睛的中心.
|
| 关 键 词： | 人眼对定位 滞后阈值法 支持向量机 投影函数 |
EYE PAIR LOCALIZATION IN COLOR IMAGES |
| |
| Gu Xiaobo,Wang Yaoming.EYE PAIR LOCALIZATION IN COLOR IMAGES[J].Computer Applications and Software,2008,25(8). |
| |
| Authors: | Gu Xiaobo Wang Yaoming |
| |
| Affiliation: | Gu Xiaobo1 Wang Yaoming21(Mathematics , Science College,Shanghai Normal University,Shanghai 200234,China) 2(Department of Information Engineering,Shanghai Dianji University,Shanghai 200240,China) |
| |
| Abstract: | In this paper,an approach for eye localization in colour images is proposed. Firstly,in order to decrease the search areas when SVM verifying eyes,the hysteresis thresholding technique is adopted to segment the skin likelihood image in YCbCr colour space modelled by Gaussian Mixture Model.Candidate areas are then further determined by using chrominance,luminance,gradient information and regional rules of eyes,false alarm regions in SVM detection are eliminated by using geometric distribution of eyes. Finall... |
| |
| Keywords: | Eye pair localization Hysteresis thresholding Support vector machines Projection function |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
