Caricature System using Moving Least Squares based on Justin

In this paper, we propose a new automatically generation system of caricature which emphasizes the unique characteristics of the input images. This system can create the cartoon character similar to user’s features by using set rules adopted a particular artist's technique. Input image is transformed by using MLS(Moving Least Squares) approximation[1] based on a predefined cartoon character’s image. The unique characteristics of user can be detected by comparing the information between the mean face feature and the input face feature extracted by AAM(Active Appearance Model) [2]. In this paper, to exaggerate the detected unique characteristics, we set-up the exaggeration rules using the technique of Justin[10] who is caricature artist. In addition, during the cartooning process, user's hairs and accessories are applied to the deformed image to make a close resemblance. Reliable and matured caricature can be represented through the exaggeration rules of the actual caricature artist's techniques. From this study, we can easily create a cartoon caricature appearing user's feature by combining a caricature with existing cartoon researches.