首页 | 本学科首页   官方微博 | 高级检索  
     

基于最小平方差模型的改进半色调技术
引用本文:王小华,刘丽娜.基于最小平方差模型的改进半色调技术[J].计算机工程与设计,2007,28(10):2381-2383.
作者姓名:王小华  刘丽娜
作者单位:杭州电子科技大学计算机学院,浙江杭州310018
摘    要:基于模型的半色调技术是根据显示设备和人眼视觉系统的特性从而使图像质量达到最优,其中基于最小平方差模型的半色调技术是其中一种比较好的方法,而基于最小平方差模型的改进半色调技术则是在前者的基础上进行了改进,提出了一种新的迭代方法即取反/交换法.该技术的基本思想是对二进制图像进行打印模型和视觉模型处理,和对原始灰度图像仅进行视觉模型处理,使两者之间的输出图像之间的误差最小.实验证明该技术比传统技术具有更高的分辨率和更多的灰度级,图像更清晰.

关 键 词:基于模型的半色调  人眼视觉系统  最小平方差模型  打印模型  视觉模型  取反/交换法  误差最小  方差模型  改进  半色调技术  灰度级  分辨率  传统技术  验证  输出  灰度图像  处理  视觉模型  二进制  思想  交换法  迭代方法  比较  最优  图像质量  特性
文章编号:1000-7024(2007)10-2381-03
修稿时间:2006-03-28

Least-squares model-based modified halftoning
WANG Xiao-hua,RIU Lina.Least-squares model-based modified halftoning[J].Computer Engineering and Design,2007,28(10):2381-2383.
Authors:WANG Xiao-hua  RIU Lina
Affiliation:Institute of Computing Technology, Hangzhou Dianzi University, Hangzhou 310018, China
Abstract:Model-based halftoning techniques exploit the properties of the display device and the human visual system to maximize the quality of the displayed images. Among these algorithms, the least-squares model-based technique is a better approach to produce reproductions. A least-squares model-based modified technique bases on the former and improves it, a new iterative technique that is toggle/swap technique is proposed. The basic idea of this technique is that it exploits both a printer model and a model for visual perception. It attempts to produce an optimal halftoned reproduction, by minimizing the squared error between the response of the cascade of the printer and visual models to the original image and the response of visual models to the original gray-scale image. Experiments show that a least-squares model-based modified halftoning produces better textures and higher spatial and gray-scale resolution than conventional techniques. Moreover, the picture is much sharper.
Keywords:model-based halftoning technique  human visual system  a least-squares model  printer model  visual model  toggle/swap technique
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号