Realistic illumination for outdoor AR environment

Augmented reality is very popular these days, and many studies are being conducted to increase the user experience by mapping the augmented reality environment to real-world conditions. In this thesis, it is aimed to illuminate the virtual object in the mobile augmented reality environment according to the current real-world conditions. At the same time, another contribution of this study is to conduct an experiment to observe how sensitive the users are to the errors that may occur in the lighting conditions on the virtual object. In order to provide realistic illumination, first of all, the Unity machine's location service and Unity's compass value were used, along with geographic calculations, to find the Sun's position during the day. In order to collect the data set for this study, an AR application was developed that will allow us to manually adjust the virtual object to the real-world object and match it, and also to obtain data from online sources. With this application, the average light intensity, hue, and saturation values calculated in the background for the image taken from the camera and the manually adjusted Unity sky box material parameters such as atmosphere thickness, sky exposure values, and Unity directional light parameter shadow sharpness values were saved in a file. Shadow sharpness, sky exposure, and atmosphere thickness values were taken as dependent values and analyzed separately by multiple linear regression method. Looking at the results of the analysis, the mean square error obtained from the function created for the sky exposure was 0.25, 0.03 for the shadow sharpness value, and 0.04 for the atmosphere thickness value. Looking at the results of the experiment, it was observed that the users were most sensitive to the light intensity value and less sensitive to the error on the shadow sharpness value.