›› 2016, Vol. 29 ›› Issue (5): 124-.

• 论文 • 上一篇    下一篇

改进的小波阈值图像降噪算法实现

甘航萍,王力   

  1. (贵州大学 大数据与信息工程学院,贵州 贵阳 550025)
  • 出版日期:2016-05-15 发布日期:2016-05-24
  • 作者简介:甘航萍(1990—),女,硕士研究生。研究方向:图像处理,数据挖掘。王力(1971—),男,教授,硕士生导师。研究方向:模式识别,信号处理与机器学习。

Improved Wavelet Threshold Image De-noising Algorithm

GAN Hangping,WANG Li   

  1. (College of Big Data and Information Engineering,Guizhou University,Guiyang 550025,China)
  • Online:2016-05-15 Published:2016-05-24

摘要:

针对图像降噪中的不连续性、存在恒定偏差等问题,在Dohono等提出的硬、软阈值法基础上,提出了一种改进的指数函数阈值降噪算法。与传统阈值降噪法相比,该方法不仅克服了传统方法的不足,且具有更好的灵活性。仿真实验结果表明,该改进阈值法的降噪效果无论在信噪比增益及均方误差上,或是在视觉效果上,均优于传统阈值降噪法。

关键词: 小波阈值降噪, 信噪比, 均方误差, 图像复杂度

Abstract:

An improved threshold de-noising exponential function is proposed on the basis of hard and soft threshold method put forward by Dohono to address the discontinuity and constant bias problems in image noise reduction.The threshold value method overcomes the deficiency of the traditional threshold method and offers good flexibility.Simulation results show that the improved threshold method of noise reduction is superior both in SNR gain and MSE and in visual effect to the traditional threshold noise reduction method.

Key words: wavelet threshold de noising;SNR;MSE;image complexity

中图分类号: 

  • TP391.14