The Research On Web-based 3D Avatar Generation And Control

With the development of virtual reality and augmented reality(AR)technology,new types of social media continue to emerge and develop rapidly.One of the most important technologies of these social media applications is to generate an avatar similar to that for users.The purpose of this article is to generate an exclusive avatar for the user,and realize that the avatar mimics the user's facial expressions at the same time,so that the avatar can become the avatar of the user in the online world.Existing avatar generation technologies are mainly divided into methods based on 3D scanning equipment and methods based on model library matching.Among them,the method based on the three-dimensional scanning device has high requirements on the device,which is not convenient to use and popularize.The method based on the model library matching method has small individual differences in results,which easily causes aesthetic fatigue.In order to solve the shortcomings of the existing methods,a texture production method based on Delaunay triangulation and affine transformation of graphics is proposed here.Combined with the 3D model generated by the face reconstruction on the web side,it can generate exclusive avatars with personal characteristics for users without relying on any three-dimensional equipment and established library of facial elements.Existing avatar expression control is mainly achieved through the muscle-bone binding of the model.This method requires the model to be set in advance and the model to be bound by the designer,which causes the model to be fixed and a lot of manual consumption.In order to solve the above problems,a facial expression animation method based on region division is proposed here.Combining gesture estimation technology and face posture normalization method,the expression control of the generated avatar model can be performed without human participation,so that the avatar can emulate user emoticons in real time.The experimental results show that the method proposed in this paper can generate personalized and exclusive three-dimensional avatars based on the user's two-dimensional face information.Combined with the control of the avatars,it can realistically and reliably simulate the user's facial actions.It provides a feasibility reference for web-based avatar related research.