Cartoon Motion Capturing and Retargeting by Rigid Shape Manipulation

Abstract

Motion capture from live performance has received a lot of attention in the past few years. However, little work has been done on capturing the motions from existing cartoons. This paper presents a novel approach for capturing motion from existing cartoons and retargeting it to new characters in order to animate them. Most current approaches rely on the identification of key shapes and they fail to directly handle articulated shapes or local deformations. In contrast, we propose to use key-points as more efficient descriptors of motion. We use shape context to capture the motion between successive source frames. Then for retargeting the captured motion to a new character, we use an efficient rigid shape manipulation method that handles local deformations. The proposed method relies on user interaction only for the first source frame. Our algorithm has been applied to a set of test cases and the results shows improved performance in preserving the target's visual style particularly for articulated objects.