计算机工程 ›› 2011, Vol. 37 ›› Issue (12): 187-189.doi: 10.3969/j.issn.1000-3428.2011.12.063

• 图形图像处理 • 上一篇    下一篇

基于二阶导数算子与小波变换的图像去噪

王绪四 a,杨恢先 a,谢鹏鹤 a,满 莎 b,彭 友 a   

  1. (湘潭大学 a. 材料与光电物理学院;b. 信息工程学院,湖南 湘潭 411105)
  • 收稿日期:2010-11-23 出版日期:2011-06-20 发布日期:2011-06-20
  • 作者简介:王绪四(1986-),男,硕士研究生,主研方向:图像处理,模式识别;杨恢先,教授;谢鹏鹤,硕士研究生;满 莎,硕士;彭 友,硕士研究生
  • 基金项目:
    海南省自然科学基金资助项目(60897);海南省教育厅基金资助项目(Hj2009-135)

Image De-noising Based on Second Derivative Operator and Wavelet Transform

WANG Xu-si a, YANG Hui-xian a, XIE Peng-he a, MAN Sha b, PENG You a   

  1. (a. College of Material and Photoelectronic Physics; b. College of Information Engineering, Xiangtan University, Xiangtan 411105, China)
  • Received:2010-11-23 Online:2011-06-20 Published:2011-06-20

摘要: 二阶导数算子噪声定位的图像去噪法对椒盐噪声有很强的去噪能力,但对高斯噪声去噪效果较差,基于小波变换的图像去噪法能有效去除高斯噪声,但几乎不能去除椒盐噪声。针对上述问题,采用二阶导数算子降噪与小波变换去噪相结合的方法对图像去噪,利用2种方法进行优势互补,能较好地去除椒盐、高斯噪声和椒盐-高斯混合噪声,降低选择阈值的难度,有利于提高图像去噪精度。实验结果表明,该算法是有效可行的。

关键词: 二阶导数算子, 椒盐噪声, 高斯噪声, 小波变换, 图像去噪

Abstract: An image de-noising method based on second derivative operator to noise location can de-noising impulse noise effectively, but it is not very good to remove Gaussian noise. An image de-noising method based on wavelet transform has the ability to remove Gaussian noise while it hardly de-noises impulse noise. An image de-noising method based on second derivative operator and wavelet transform has some advantages. It has better effect on removing impulse noise, Gaussian noise and impulse-Gaussian mixed noise and it decreases the difficulty of threshold selection, which is good to improve the accuracy of image de-noising. Experimental result shows that this method is effective and feasible.

Key words: second derivative operator, impulse noise, Gaussian noise, wavelet transform, image de-noising

中图分类号: