A Comparative Analysis of Augmented Reality Frameworks Aimed at Diverse Computing Applications
Mfundo Andrew Maneli, Omowunmi Elizabeth Isafiade · 2022
Immersive systems such as Augmented Reality (AR) and Virtual Reality (VR) have proven useful in diverse computing domains. However, there is little effort on accuracy measurements within AR applications, which could greatly impact outcomes and decisions in certain domains, such as crime scene investigations, among others. This paper aims to analyze and evaluate two existing prominent AR frameworks, ARCore and ARKit, which support the development of diverse mobile computing applications for immersive systems. This research developed prototype applications and conducted comparison tests of measurement accuracy within the applications. The accuracy was tested using four distance criteria across six different devices, spanning ARCore and ARKit frameworks. A control experiment was used to benchmark the measurement accuracy. Relatively, an instance of the experiment presented ARCore as reliable. Overall, ARKit proved to be more accurate between the two frameworks, with an average accuracy of 99.36% as opposed to 89.42% scored by ARCore. The obtained results can give insight on the choice of framework to consider during AR application development for a specific domain, hence boosting quality of experience.