摘要
为了让普通用户方便地得到像素画卡通头像,提出一种自动从人脸照片生成像素画的算法.首先通过对脸部特征定位找到关键特征,然后在低分辨率下对特征的形状和位置进行优化调整;在调整后的各个区域采用"画家调色盘"色彩校正方法对其颜色进行校正,生成像素画.实验结果表明,该算法能够生成结构清晰、色彩明快的像素画.
In this paper, we introduce an automatic facial pixel drawing generation algorithm for common users. Firstly, we locate the key facial feature, and optimize the location and size of each feature by local adjustment. Colors in each adjusted region are then processed using the proposed “artist color palette” method. Experimental results show that our method can generate pixel drawing style cartoon with recognizable feature in low resolution display.
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2009年第12期1733-1739,共7页
Journal of Computer-Aided Design & Computer Graphics
