基于图像的个性化卡通系统

介绍了一个卡通动画系统.该系统可使用户很方便地生成个性化的卡通头像,并通过简单的交互定制各种夸张而生动的表情,还可以实时生成语音驱动的卡通动画.系统主要包括3部分:基于图像的卡通自动生成模块、交互的卡通表情编辑模块和语音驱动的卡通动画模块.卡通自动生成模块采用基于样本学习的方法生成具有特定艺术风格的卡通头像.这里使用了一种非参数化采样算法来学习原始照片和画像间复杂的统计关系.在此基础上,卡通表情编辑器可以通过预先设计的模板生成各种具有生动表情的卡通.同时,用户也可以定制自己的表情模板.卡通动画模块使用了实时的唇形合成算法,可以生成语音驱动的卡通动画.该系统可以使普通用户很方便地生成个性化的卡通头像和动画,在这一点上优于其他动画制作系统.

A Personalized Image-Based Cartoon System

In this paper, a Cartoon system is presented, which can generate a personalized Cartoon face from an input image. The system is designed for general users. After a realistic-looking Cartoon face automatically generated by this system, attractive Cartoon expressions and animation can be created with little user-interactions. The system consists of three main components, an image-based automatic Cartoon generator, a template-based Cartoon editor, and a speech-driven Cartoon animator. An example-based approach is taken by Cartoon generator to generate the Cartoon face while capturing an artist style.An inhomogenous non-parmetric sampling is employed to lwarn the atatistical relationship between the original image and the drawing.For editing,pre-designed expression templates are supplied by Cartoon editor editor.A real time lip-syncing algorithm is also developed to make Cartoon animation easily.The Cartoon and animation generated by this system have personal style.This is the main advantage of this system over others.