Finding Hidden Shrines using AR and Clustering Techniques

Abstract

This paper describes the use of augmented reality technology in an offline mobile application to recognize and find shrines. This application uses clustering algorithms to cluster shrines and then a convex hull is used to combine the geographical coordinates of multiple shrines. The photographs and videos of shrines can be downloaded by tapping on any clusters. Images downloaded will have their features extracted and stored in an augmented database which then will be used by ARCore, which is an augmented reality API, to identify and recognize all shrines in that cluster.