The special make-up composition system by estimation of virtual backbone

Abstract

In this paper, we have proposed a facial image entertainment system, which can assume virtual backbone and virtual fat of user or modify the parts of user. To define virtual backbone, we used snake and template matching methods to detect the contour of user, and other parts, like eyes, nose and so on. And we use EM method to make a simulation of virtual fat. With the defined virtual backbone, system modifies 3D model, and map the special make-up to compose the output image. User can modify image, so that user can make an image without any disharmony. This proposal can handle many kinds of situation. E.g. different face colors, face shapes, view of cameras, proposed system can handle stably.