Avatar Generation Using Facial Component Matching

Nowadays, avatars are widely used in games and Internet environments. Especially, video game consoles such as Wii (Nintendo) use avatars for representing the user's alter ego. Many researchers have suggested automatic avatar generation methods, most of which generate avatars by simplifying images using non-photorealistic rendering techniques. In this paper, we suggest an example-based method for generating avatars. Our method generates an avatar which is similar to the input front-view photograph by matching each facial feature to the corresponding avatar features. The system first extracts facial features such as eyes, eyebrows, a mouth, a nose, and a face outline, from the input image. Then it matches them to the corresponding avatar components using the graph similarity and the Hausdorff distance. Finally, the system generates an avatar by compositing the most similar components.