基于特征流的抽象线条画绘制

CORC > 软件研究所 > 软件所图书馆 > 期刊论文

	基于特征流的抽象线条画绘制
	王山东 ; 刘学慧 ; 陈彦云 ; 吴恩华
刊名	计算机学报
	2014
卷号	37 期号:3 页码:611-620
关键词	非真实感绘制线条画绘制边缘切向流各向异性滤波
ISSN号	2544164
其他题名	Abstract line drawings from photographs using flow-based filters
通讯作者	Wang, S.-D.(sdwang10@gmail.com)
中文摘要	为了将照片图像转换为具有一定艺术美感的线条画图像,文中设计了3种基于特征流的各向异性滤波器:FGsD滤波器、FGaD滤波器和FLSM 滤波器.这些滤波器的主要任务是提取图像的边缘信息,并将其显示为光滑连续的风格化线条.前两种滤波器是在分析数字图像中基于一阶微分和二阶微分边缘检测算法的性能后,将高斯一阶导滤波结果和高斯差分滤波结果进行适当的混合作为边缘检测的微分响应,然后对混合的微分响应值进行柔和阈值化处理提取边缘点.第 3种滤波器是专门针对图像中的线型边结构提出的,通过计算局部亮度相似度质量来判断该像素是否属于边缘点.如果将FGsD滤波结果和FLSM 滤波结果进行叠加,还可以得到明暗对比度增强的抽象线条画效果.与已有的线条画绘制算法相比,采用文中算法所生成的线条画视觉特征更鲜明、风格化效果更突出、艺术表现力更强烈.
英文摘要	This paper proposes three flow-based anisotropic filters for stylizing a photograph in the line drawing style: FGsD filter, FGaD filter and FLSM filter. The main tasks of these filters are to detect edges and display them with a set of clean, smooth, coherent and stylistic lines. Based on the observation that edges are extracted and localized by finding extreme values of the first derivative or zero crossings of the second derivative of luminance function, the first two filters are defined by properly combining the first Gaussian derivative and the second DoG operator for detecting edges with smoothed step function. The third filter uses the local luminance similarity mass to specially detect line edges in the image. Blending the FGsD and FLSM filtering results is particularly useful for enhancing features and conveying a hand-painting style that can elicit an esthetic response from viewers. Experimental results show that our approach can produce more attractive and impressive line illustrations with a variety of photographs.
收录类别	EI ; CSCD
语种	中文
CSCD记录号	CSCD:5064071
公开日期	2014-12-16
内容类型	期刊论文
源URL	[http://ir.iscas.ac.cn/handle/311060/16765]
专题	软件研究所_软件所图书馆_期刊论文
推荐引用方式 GB/T 7714	王山东,刘学慧,陈彦云,等. 基于特征流的抽象线条画绘制[J]. 计算机学报,2014,37(3):611-620.
APA	王山东,刘学慧,陈彦云,&吴恩华.(2014).基于特征流的抽象线条画绘制.计算机学报,37(3),611-620.
MLA	王山东,et al."基于特征流的抽象线条画绘制".计算机学报 37.3(2014):611-620.