Generation of virtual images under different lighting conditions*

摘要

In this paper, a new method of generating virtual images of the same scene under different lighting conditions is presented, and the inherent relations of color values and outside lighting conditions are discovered by the use of general regression neural networks. As a result, novel images of same scene under different lighting conditions can be deduced from known scene image. The method can infer the change of lighting in the scene without any space geometry informations. Dynamic reconstruction of scene can be reached combining IBMR techniques.