Mobile augmented reality tracking registration method based on online study

Abstract

The invention provides a mobile augmented reality tracking registration method based on online study. The method includes the steps that back projection errors are reduced to optimize a position matrix of a camera, a coordinate system of a virtual three-dimensional object and a coordinate system of a scene plane are built, and then registration of the virtual object to a real scene can be more accurately finished; lost feature points are restored through a local region feature point fast searching algorithm, the detected scene feature points are online studied, weights of scene three-dimensional point coordinates are updated in real time, corresponding weights of three-dimensional points with large contributions are increased step by step, corresponding weights of three-dimensional points with small contributions are decreased step by step, a global value maximum projection matrix is calculated through study, and therefore augmented reality registration accuracy is further improved.

Description

technical field [0001] The invention belongs to the technical field of mobile augmented reality, and in particular relates to a mobile augmented reality tracking registration method based on online learning. Background technique [0002] Augmented Reality (AR) is one of the current research hotspots in the field of computer vision. It enhances people's understanding and experience of the real environment by closely combining virtual objects with the real environment. The characteristics of augmented reality technology are: combination of virtual and real, real-time interaction and tracking registration. Among them, tracking registration is one of the core problems that augmented reality technology needs to solve. In particular, the vision-based marker-free tracking and registration technology has gradually become a research hotspot in the field of augmented reality due to its high precision, simple equipment and no interference to the environment. [0003] Most of the trad...

AI Technical Summary

Problems solved by technology

[0005] In 2011, scholar T.Guan et al. (document 1T.Guan; L.Y.Duan. "Real Time Camera Pose Estimation for WideArea Augmented Reality Application". IE EE Computer Graphics and Application, 31(3): 56-68, 2011) was able to correctly identify For natural scenes, a large number of reference views are saved, and natural scenes are recognized through local feature matching methods, but the algorithm is very computationally intensive and cannot be applied to smartphones

However, it is necessary to use a computer to perform 3D reconstruction of all scenes in advance, to establish the 2D relationship between the 3D points of the scene and the related images of the scene, and to manually download the coordinates of the 3D points of the scene where the user is located, in order to realize the subsequent tracking and registration algorithm. Otherwise, a huge search space will be generated, which is difficult to meet the real-time requirements

Images