Intelligent Marker Repository for Augmented reality based Autonomous Navigation

Abstract

This paper demonstrates the concept of self-created and self-refined repository for the augmented reality based autonomous navigation system. The repository comprises of human understandable markers detected during the bare run for an unknown environment. These markers along with the navigation instructions will help the system to navigate autonomously on the road. The novelty of the work lies in the detection of three types of markersfrom the run time video to support navigation task and categorizing them as main markers, supportive markers and backup markers. These markers are also used in the next runs for path validation. Main markers are the most prominent markers and in the absence of the main markers, supportive and backup markers help in path validation. After storing these markers, this self-created repository refines itself during the next runs on the basis of marker usability.MATLAB tool and NoSQL database have been used to get the desired results.