Comparative study of the performance of real-time inverse lighting with matte, semi-gloss and gloss surfaces

Abstract

Augmented Reality (AR) is the interactive process of introducing virtual objects or characters to real-world scenes. An effective way to increase the realism in AR is by mimicking real-world lighting conditions on the virtual objects. The process of gathering and analyzing real-world lighting information is called inverse-lighting. The surface textures of real-world objects may have different levels of glossiness. The goal of this research is to compare the effects that different glossiness levels have on the outcomes of the calculations. Several models of a regular dodecahedron were created using the Blender modeling software. These models were used to calculate and compare inverse-lighting on different levels of surface glossiness. Physical dodecahedrons also were created and used to see whether the Blender models accurately represent reality.