The Design And Implementation Of 3D Face Reconstruction System

Nowadays,the way of our life is affected by high technologies.In the field of computer vision,3D reconstruction has always been one of the hottest research directions.With the continuous advancements of computer technology,3D modeling is gaining wider and wider applications due to its own advantages.Human face is special because it is the most direct and important carrier for human emotion expression and communication.And with the popularization of smart phones,mobile phones have become an indispensable item for people to travel and gradually occupy an important position in life,so it is very meaningful to combine them.Thus,this work design and implement a human face reconstruction system based on Android.In this paper,a 3D face reconstruction algorithm using 3DMM and an improved 3DMM by correcting residual network is designed and this algorithm implements a 3D face reconstruction system with an Android mobile phone as client and a host as server.This article first implements an image acquisition method on an Android phone to obtain a 2D face image,and then uses the open source OkHttp library to communicate between the client and the web server flask.Then the host uses the configured flask to start 3D face reconstruction.After the reconstruction,it uses OkHttp request data from the flask and transmits them back to the client.Finally,the client is using OpenGL ES for 3D face painting.This paper adopts the C/S architecture and realizes the data transmission between the client and the host server through network communication.Our method can be represented timely under local networks.Finally,by comparing the reconstruction effect on the server side,we can see that the improved 3DMM reconstruction effect using the improved residual network is using more detailed information compared with original 3DMM.Using this method,this work finally achieves a system that can reconstruct a 3D face model in real time after taking a photo on an Android phone.