肖像照片漫画风格化方法研究

Research on Stylization Method of Portrait Cartoon

本文针对相关方面的问题开展了分析,围绕数据驱动及滤波方法对线条画智能生成方法进行了讨论。具体来看,分别从五官与头发两方面进行了分析,借助数据及非数据驱动方式,依次对五官及头发合成进行了讨论,主要是基于已有特征点之上为其匹配相应的数据,从而生成五官图片;对于头发的处理,本文讨论了多种处理方式,主要是生成线条域,为实现最佳效果,还讨论了纹理和线条域叠加的方式,以此生成多样风格的头发。将五官及头发图片进行整合,将形成漫画风格的线条画。

This paper analyzes the related problems,and discusses the intelligent generation method of line drawing around the data-driven and filtering methods.Specifically,from the facial features and hair two aspects analysis,with data and non-data-driven way,in turn to the facial features and hair synthesis is discussed,mainly based on the existing features to match the corresponding data,to generate the facial picture of hair,this paper discusses a variety of processing,mainly to generate line domain,in order to achieve the best effect,also discussed the superposition of texture and line domain,in order to generate various styles of hair.Integrating the facial features and hair pictures will form a comic-style line painting.