A Sensor Based Approach to Outdoor Illumination Estimation for Augmented Reality Applications on Mobile Devices

Abstract

Realistic Augmented Reality applications require integrating virtual objects in the real world in a seamless visual way. The main problem to obtain a perfect visual augmentation is rendering the virtual objects with consistent illumination. In this paper, we present a novel approach to estimate outdoor illumination for Augmented Reality applications on mobile devices based on information acquired from sensors. We use an ambient light sensor to automatically detect in real-time the scene illuminance, which is used to simulate the actual lighting conditions. As outdoor illumination is mostly dependent on the weather, we derived the sky/sun illumination parameters from the light sensor data based on the IESNA sky model. The sunlight direction is estimated based on information from the GPS, date and time of the day. The result is a practical solution for outdoor illumination estimation handling light changes under natural conditions and applying them to the rendering of virtual objects in the real scene.