Monocular Visual SLAM for Markerless Tracking Algorithm to Augmented Reality

Augmented Reality (AR) tries to seamlessly integrate virtual content into the real world of user. Ideally, would behave exactly like objects. This necessitates a correct and precise estimation user’s viewpoint (or that camera) with regard content’s coordinate system. Therefore, real-time establishment 3-dimension (3D) maps in scenes is particularly important for augmented reality technology. So this paper, we Simultaneous Localization Mapping (SLAM) technology reality. Our research implement an system without markers using ORB-SLAM2 framework algorithm. In paper propose improved method Oriented FAST Rotated BRIEF (ORB) feature extraction optimized key frame selection, as well use Progressive Sample Consensus (PROSAC) algorithm planar implementations, thus solving problem increased runtime because loss large amounts texture information images. get better results by comparing experiments data analysis. However, there are some methods PROSAC which more suitable detection plane points.