3D Personalized Face Modeling Based on KINECT2
Jia Li, Yue Zhang, Pin Xu, Shanzhen Lan, Shaobin Li · 2016
As a challenging topic in computer graphics and computer vision, 3D face modeling has been applied in various fields including film and animation production, game development as well as medical analysis. In this paper, we mainly aims at face personalized modeling optimization based on KINECT2. The optimization of the scheme proposed is the mesh deformation based on the differential coordinates constrained by face++ feature points and facial muscle function, namely firstly by extracting the location of feature points on the color image and depth map to obtain the feature information of the personalized face, then aligning the intermediate results of the KINECT2 modeling to the feature information. It is good to keep the local grid differential property, at the same time. In this paper, the automatic UV mapping is realized by means of planar differential mesh deformation, which improves the accuracy and efficiency of manual adjustment of UV coordinates in Maya. The experimental results show that the optimization results are more fit in with the individual face.