Research And Implementation Of 3D Model Replacement Technology Based On AR Real Scene

Augmented Reality(AR)is an experience of users interacting with the real world environment and is a hot research field in recent years.At the same time,the user is allowed to interact with virtual things and information.Virtual things and information,as a complement to the real scene,achieve the effect of enhancing reality.In general,it is the use of computers to add some computer-generated things when displaying actual scenes,so that people can feel things and scenes more vividly and intuitively.As more and more AR SDKs start to appear,such as Vuforia,EasyAR and so on.Currently,and most of them are based on image recognition tracking and then rendering display presets on them.Good content.This paper is based on the existing AR system.Based on the image recognition and tracking technology,the real 3D object is tracked and recognized on the mobile device and its corresponding virtual 3D model is rendered.The user can interact in real time and expand the AR on the mobile device.Application area.Firstly,a two-dimensional image sampling feature point algorithm ORB is used to establish a three-dimensional space coordinate.In this space coordinate system,the spatial feature points of the three-dimensional object are sampled.Then the three-dimensional object is tracked.The three-dimensional tracking algorithm in this paper adopts the PTAM algorithm.The framing method has been modified to improve the tracking accuracy of 3D objects,and the performance on mobile devices has been improved accordingly.Then 3D Max modeling software is used to create virtual 3D objects with the same surface texture.After recognition,the rendering is performed;finally,the user can interact with the three-dimensional object rendered after the recognition is successful.