作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2011, Vol. 37 ›› Issue (24): 10-12. doi: 10.3969/j.issn.1000-3428.2011.24.004

• 博士论文 • 上一篇    下一篇

基于HSI彩色空间模型的域替换降噪算法

代 煜,周海锋,王 萌,张 生,金 纬   

  1. (南开大学信息技术科学学院,天津 300071)
  • 收稿日期:2011-09-28 出版日期:2011-12-20 发布日期:2011-12-20
  • 作者简介:代 煜(1981-),男,讲师、博士,主研方向:医学图像处理,医疗机器人控制与智能化,微弱信号检测,立体视觉技术;周海锋、王 萌、张 生、金 纬,本科生
  • 基金资助:

    国家“863”计划基金资助项目(2009AA044001);国家创新性实验计划基金资助项目(101005513)

Denoising Algorithm of Domain Replacement Based on HSI Color Space Model

DAI Yu, ZHOU Hai-feng, WANG Meng, ZHANG Sheng, JIN Wei   

  1. (College of Information Technical Science, Nankai University, Tianjin 300071, China)
  • Received:2011-09-28 Online:2011-12-20 Published:2011-12-20

摘要: 对彩色数字图像进行降噪处理时,为兼顾标量处理法的简便性和矢量处理法的有效性,提出一种对彩色空间域实行替换的方法。采用均值滤波法在RGB域内进行滤波处理,将图像数据从RGB域转换到HSI域,用滤波处理后的H域、S域替换含有噪声的H域、 S域。对由双目内窥镜采集的实际彩色图像进行实验,结果证明该域替换法可以结合标量处理和矢量处理的优点,达到比一般标量处理法及简单矢量处理法更好的降噪效果。

关键词: HSI彩色空间, 域替换, 降噪算法, 低信噪比, 标量矢量结合, 分量处理

Abstract: Considering the simplicity of scalar processing and the effectiveness of vector processing when doing color digital image noise reduction, a method of domain replacement is proposed in this paper. It processes the image data in RGB space with mean filtering, converts the image data from RGB to HSI, and replaces the original H and S domain with that filtered. The noise in H and S domain is reduced indirectly after these operations. The method is tested with an actual color image captured from a binocular endoscope. According to the results of the test, the method of domain replacement may have both the advantages of scalar processing and vector processing. Compared with common scalar processing and simple vector processing, the method can achieve a better effect.

Key words: HSI color space, domain replacement, denoising algorithm, low Signal Noise Ratio(SNR), combination of scalar and vector, component processing

中图分类号: