以计算机人脸漫画生成的流程为主线,总结了近年来计算机生成人脸漫画技术的国内外研究现状,将现有的方法分为基于高层语义描述和基于图像模板,介绍2种方法生成人脸漫画的流程.根据夸张规则的不同,介绍了基于人机交互、基于样本模板、基于绘画规则以及基于统计学习4种肖像漫画生成方法,并对各方法所涉及的核心算法进行对比分析.结合近年来的研究热点,给出了人脸漫画未来的研究难点和发展趋势.%An overview of the development in the field of computer caricature generation is presented based on the process of creating caricature.The established approaches are based on high-level semantics and image template,which.Then four approaches are introduced respectively according to the different exaggerated rules.Their advantages and limitations are illustrated in detail.Finally,the technical challenges and future developments are pointed out.
展开▼