Subjective and Objective Quality Assessment for Augmented Reality Images

Extended Reality (XR), including Virtual Reality (VR), Augmented Reality (AR), and Mixed Reality (MR), has rapidly advanced, offering immersive experiences to users. Among these technologies, AR has gained prominence for seamlessly integrating virtual and real-world scenes, enabling the augmented experience. However, assessing the quality of AR images remains a critical challenge for the lack of investigation on the interaction between the virtual-world contents and real-world scenes in real cases. In this work, we construct an AR image quality assessment database in real applications based on an AR HMD. Then we analyze the subjective quality ratings in real AR environments and perform objective quality assessment using state-of-the-art models. This comprehensive subjective and objective analysis enhance our understanding of the Quality of Experience (QoE) in AR.