Two Stage Fuzzy Inference Systems for Autonomous Lighting in 3D Animated Movie Scene

Abstract

Lighting is one of the very important things in cinematography. Lighting in cinematography is used not only to illuminate the objects, but it has many important roles such as to make the object looks sharper, has tone, depth and separation, and so on. There are some parameters that affect the lighting result, such as light source type that is used, its colour, intensity, position, and direction. Unfortunately, lighting has so many permutations and variations and there is more than one right way to light the scene. In this research, a designer designs some scenes that will be used as inputs. In this study, the authors found that lighting set is strongly influenced by five important things, which are: the situation of the scene, Point-of-View, the intensity of the attached light sources, camera position, and emotion of the characters, that we called as the SPICE system that is the abbreviation from situation, PoV, intensity, camera, and emotion. some additional light sources will be added into the scene automatically according to the scenario and director's behaviour using the SPICE system that we propose. The SPICE framework has two fuzzy inference systems in calculating the intensity of light that hits an object and some additional light sources that are added to the scene automatically to make it looks much better. This research will make a calculation to gain the best scenes as an output. There are 20 scenes used in this research. As the final result, designers assess 80% outputs have been as expected.