Objective Quality Evaluation of an Angularly-Continuous Light-Field Format

With the rapid advances in light-field display and camera technology, research in light-field content creation, visualization, coding and quality assessment has gained momentum. While light-field cameras are already available to the consumer, light-field displays need to overcome several obstacles in order to become a commonplace. One of these challenges is the unavailability of a light-field visualization format which can be used across various light-field displays. Existing light-field representations are optimized for specific displays and converting them for visualization on a different display is a computationally expensive operation, often resulting in the degradation of perceptual quality. To this end, an intermediate, display-independent and angularly-continuous light-field representation format has been proposed recently, targeted towards large-field-of-view light-field displays. In this paper, we evaluate the said data format in terms of degradation in objective quality under three compression methods. We found that, while offering display-independence, the intermediate light-field format maintains the same objective quality in general and achieves higher objective quality in some cases compared to the conventional linear camera representation