基于特征流的抽象线条画绘制  

Abstract Line Drawings from Photographs Using Flow-Based Filters

作　　者：王山东[1,2,3] 刘学慧[1] 陈彦云[1] 吴恩华[1,2,3] 

机构地区：[1]中国科学院软件研究所计算机科学国家重点实验室,北京100190 [2]澳门大学科技学院电脑与资讯科学系 [3]中国科学院大学,北京100190

出　　处：《计算机学报》2014年第3期611-620,共10页Chinese Journal of Computers

基　　金：国家"九七三"重点基础研究发展规划项目基金(2009CB320802);澳门大学研究基金资助~~

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.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.

关 键 词：非真实感绘制 线条画绘制 边缘切向流 各向异性滤波中图法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]