CASIA OpenIR  > 毕业生  > 硕士学位论文
Thesis Advisor谭铁牛
Degree Grantor中国科学院自动化研究所
Place of Conferral中国科学院自动化研究所
Degree Discipline模式识别与智能系统
Keyword人脸检测 图像分割 模板匹配 Face Detection Image Segmentation Template Matching
Abstract人脸检测是图像分割的一个经典问题,它在人的身份鉴定(人脸识别)、 表情识别、跟踪定位、视觉监控等领域有着广泛的应用。随着当今科学技术的 发展,人脸检测作为计算机视觉的一个主要应用领域,在科学技术、工业等方 面的应用必将越来越广泛。 本文对人脸检测的几个步骤分别进行了研究:在图像的预处理上,结合了 一种基于能量函数的方法进行边缘的连接和噪声的去除;并采用了一种快速算 法去除图像中的直线段;在算法的核心上本文利用形状特性来获取人脸的全局 信息,提出了一个基于梯度信息的椭圆环模板,从而能有效地应用于含有侧脸、 头部旋转等情况的复杂背景图像中;为了提高检测的速度,存算法的丌始对输 入图像采用了降低分辨率技术;最后我们用两个有代表性的人脸数据库以对算 法进行检验。同传统的同类算法相比,我们提出的算法有了很大的改进,对人 眼的视觉功能进行了一定程度的模拟,在对人脸位置变化的鲁棒性以及正确率 上都取得了很好的效果。
Other AbstractFace detection is one of the classical problems of image segmentation. It has wide applications in areas such as personal identification (face recognition), gesture recognition, people tracking, visual surveillance and so on. As one of the important research directions of Computer Vision, face detection has received a great deal of attention from researchers worldwide. This dissertation addresses a number of important issues in face detection. An energy function is used for edge linking and noise removal in the preprocessing step; an edge direction based template is proposed to model the face contour; then our method can be effectively used for images with various pattern faces and complex backgrounds. For the sake of computational efficiency, multi-resolution processing is used before face detection. Our algorithm is tested on two representative face databases to verify the effectiveness. Compared with conventional methods, our system is found to be more robust to noise and head pose variations.
Other Identifier502
Document Type学位论文
Recommended Citation
GB/T 7714
王建国. 人脸检测[D]. 中国科学院自动化研究所. 中国科学院自动化研究所,1999.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[王建国]'s Articles
Baidu academic
Similar articles in Baidu academic
[王建国]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[王建国]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.