摘要
为了简单、便捷、快速地实现人物肖像漫画创作,设计一种基于安卓平台的自动化肖像漫画生成系统。系统采用主动形状模型方法,对大量真实人脸进行特征分析得到平均人脸特征库,并对用户输入的人脸肖像进行特征检测、提取与筛选,然后结合漫画艺术家对人脸漫画的作画规律,对筛选的特征以及与平均人脸差异较大的特征进行归类。对不同特征采用不同夸张化方法,配以不同夸张尺度,再由计算机完成自动形变,以实现人脸肖像漫画的自动艺术化夸张,得到接近漫画创作者水平的人脸漫画。通过该系统可大幅减少漫画工作者的工作量,使普通用户也能便捷地进行人脸肖像漫画创作。
In order to get a simple,convenient and rapid way to achieve the portrait caricature creation,this system designs an auto. mated portrait caricature generation method on Android platform. By using the active shape model,we analyze a large number of real faces to get the average facial features. By characterization,extraction and filtering of the portrait for the importing of the user,the char. acteristics of the real face and the average faces’difference are classified according to the painting laws of the comic artist. For each of the different features is used in different exaggeration methods,with different exaggeration scale by the computer to complete the defor. mation automatically,so we can achieve the face of the comic cartoon art of exaggeration which is close to the artist-level portrait cari. catures.The portrait caricatures automatic production technology of the system achieves a significant reduction in the workload of comic books,and ordinary users can use the system to create portrait caricatures easily.
作者
储昭兵
王天宝
毛水凌
CHU Zhao-bin;WANG Tian-bao;MAO Shui-ling(Shanghai WuLing Shengtong Information Technology Co.,Ltd,Shanghai 200063,China)
出处
《软件导刊》
2019年第4期36-40,45,共6页
Software Guide
关键词
人脸识别
人脸漫画
肖像生成
特征提取
图像处理
face recognition
face caricature
portrait generation
feature extraction
image processing