华南理工大学学报(自然科学版) ›› 2016, Vol. 44 ›› Issue (9): 41-46.doi: 10.3969/j.issn.1000-565X.2016.09.006

• 计算机科学与技术 • 上一篇    下一篇

基于纹理与色调感知的变系数误差扩散算法

王欣波 胡建华 王云宽 吴少泓 陆浩   

  1. 中国科学院 自动化研究所,北京 100190
  • 收稿日期:2016-02-21 修回日期:2016-07-06 出版日期:2016-09-25 发布日期:2016-08-21
  • 通信作者: 王欣波( 1988-) ,男,博士生,主要从事数字印刷、图像处理研究. E-mail:xinbo.wang@ia.ac.cn
  • 作者简介:王欣波( 1988-) ,男,博士生,主要从事数字印刷、图像处理研究.
  • 基金资助:
    国家自然科学基金资助项目( 61273280) ; 中国科学院先导科技专项( XDA09020302)

An Error Diffusion Algorithm with Variable Coefficients Based on Texture Structure and Tone Awareness

WANG Xin-bo HU Jian-hua WANG Yun-kuan WU Shao-hong LU Hao   

  1. Institute of Automation,Chinese Academy of Science,Beijing 100190,China
  • Received:2016-02-21 Revised:2016-07-06 Online:2016-09-25 Published:2016-08-21
  • Contact: 王欣波( 1988-) ,男,博士生,主要从事数字印刷、图像处理研究. E-mail:xinbo.wang@ia.ac.cn
  • About author:王欣波( 1988-) ,男,博士生,主要从事数字印刷、图像处理研究.
  • Supported by:
    Supported by the National Natural Science Foundation of China( 61273280)

摘要: 传统的误差扩散算法存在非期望的人工结构以及细节表现能力不足等缺点,为此,文中提出了一种基于纹理与色调感知的变系数误差扩散算法. 首先根据图像局部视觉
偏差,在图像的平坦区和纹理区分别采用噪声调制和纹理信息进行阈值修正,然后结合蓝噪声的频谱特性以及图像的空间特性,采用蚁群算法获得各灰度级下最优的误差扩散系数. 实验结果表明,文中算法输出的半色调图像减少了误差扩散中存在的人工结构,保证了输出图像的纹理细节,具有较好的信噪比和结构相似度.

关键词: 误差扩散, 图像纹理, 噪声调制, 蚁群算法

Abstract: The traditional error diffusion algorithm may cause unexpected artifacts to produce and it can't present the details very well.In order to solve these problems,an error diffusion algorithm with variable coefficients is proposed on the basis of texture structure and tone awareness.In the algorithm,first,according to local visual errors of an image,the threshold is modified by performing the noise modulation in smooth regions and by utilizing the texture information in textured regions.Then,by taking into consideration the spectral characteristic of blue noise and the spatial characteristic of the image,the adaptive error diffusion coefficients at each gray level are obtained by using the ant colony algorithm.Experimental results show that the proposed algorithm can output halftone images to reduce the unexpected artifacts and can ensure the details of texture structure,and that it has a better performance in terms of PSNR and MSSIM.

Key words: error diffusion, image texture, noise modulation, ant colony optimization

中图分类号: