International Journal on Advanced Science, Engineering and Information Technology, Vol. 8 (2018) No. 4-2: Special Issue on Empowering the Nation via 4IR (The Fourth Industrial Revolution)., pages: 1672-1678, Chief Editor: Khairuddin Omar | Editorial Boards : Shahnorbanun Sahran Hassan, Nor Samsiah Sani, Heuiseok Lim & Danial Hoosyar, DOI:10.18517/ijaseit.8.4-2.6810

An Efficient and Robust Mobile Augmented Reality Application

Siok Yee Tan, Haslina Arshad, Azizi Abdullah


AR technology is perceived to be evolved from the bases of Virtual Reality (VR) technology. The ultimate goal of AR is to provide better management and ubiquitous access to information by using seamless techniques in which the interactive real world is combined with an interactive computer-generated world in one coherent environment. The direction of research in the field of AR has been shifted from traditional Desktop based mediums to the mobile devices such as the smartphones. However, image recognition on smartphones enforces many restrictions and challenges in the form of efficiency and robustness which are the general performance measurement of image recognition. Smart phones have limited processing capabilities as compared to the PC platform, hence the process of mobile AR application development and use of image recognition algorithm need to be emphasised. The processes of mobile AR application development include detection, description and matching. All the processes and algorithms need to be carefully selected in order to create an efficient and robust mobile AR application. The algorithm used in this work for detection, description and matching are AGAST, FREAK and Hamming distance respectively. The computation time, robustness towards rotation, scale and brightness are evaluated. The dataset used to evaluate the mobile AR application is the benchmark dataset; Mikolajczyk. The results showed that the mobile AR application is efficient with a computation time of 29.1ms. The robustness towards scale, rotation and brightness changes of the mobile AR application also obtained high accuracy which is 89.76%, 87.71% and 83.87% respectively. Hence, combination of algorithm AGAST, FREAK and Hamming distance are suitable to create an efficient and robust mobile AR application.


Augmented Reality; Image Recognition Algorithm; Detection; Description; Matching;

Viewed: 372 times (since abstract online)

cite this paper     download