基于NURBS变形和单视角图片的人脸表情生成

Facial Expression Generation Based on NURBS Warping and Single Perspective Images

传统人脸表情重建方法大多通过线性组合的方式,难以生成精细的人脸表情,并且需要使用三维激光扫描仪得到图片,不利于推广。为此,提出一种基于非均匀有理B样条建立三维人脸表情模型的方法。根据人脸形状上与表情相关的特征点调节曲面的控制点,引入几何约束和平滑约束保证整体和局部平滑。实验结果表明,与线性方法相比,该方法鲁棒性较好,可以方便地从标准图片和无约束图片中生成较为精细的三维人脸表情。

In facial expression generation,most of traditional methods are difficult to produce fine facial expression by linear combination, and the images need to be obtained with 3 D laser scanner,which is not conductive to the promotion of the method. In order to solve this problem,a method Non-uniform Rational B-spline（ NURBS） building 3 D face expression model is proposed. In this method,the control points of the surface are adjusted according to the feature points related to facial expression, and geometric constraint and smoothness constrain are to introduced to guarantee global and local smoothness. Experimental results show that compared with the linear method,this method is more robust, can easily generate fine 3 D facial expression from standard image and unconstrained image.