摘要
该文提出了一种基于单视图或小样本的多姿态人脸图像生成技术,它首先利用一个特征点集表示人脸,然后基于二元高次多项式函数最小二乘方法对人脸各姿态之间的特征点集坐标变化进行拟合,形成全局的变形域,最后由单视图通过变形映射生成多姿态人脸图像。实验结果表明,利用单视图和生成的多姿态图像进行多姿态人脸识别,正确率得到大大提高,证明该文人脸图像生成技术十分有效。
An algorithm that can synthesize multi-pose face images from a single view or small samples is proposed in this paper. First, the human face is represented with a dominant point set. Then, the variance of the dominant point set between different poses is fitted based on a least square fitting with a polynomial function and a global morphing field is formed. Finally, multi-pose face images are synthesized by image warping from a single view based on the global morphing field. An experiment of multi-pose face recognition is done based on the single view and synthesized multi-pose images. The results show that its performance is by far superior to those of the traditional methods. Thus, they prove the effectiveness of the algorithm in this paper.
出处
《电子与信息学报》
EI
CSCD
北大核心
2003年第3期300-305,共6页
Journal of Electronics & Information Technology
