부산대학교 도서관

Animated characters are the core elements in animations, possessing distinct characteristics that define the entire animation. While there has been extensive research on natural images in recent years, the study of feature extraction for animated characters remains limited compared to natural images. This paper focuses on the study of animated character portraits and aims to extract the structural and visual features of animated portraits. Different methods, such as rotation and mirror symmetry, are used to construct datasets, and an effective feature extraction method for animated character display is proposed. The validity of the algorithm is verified through experiments.