摘要
介绍了传统小波阈值去噪方法的去噪原理。小波阈值去噪的关键是阈值的设置和阈值函数的选择。该文对常规的软阈值和硬阈值函数在常规的图像去噪中存在的缺陷进行了分析,在此基础上提出了一种改进的阈值函数,改进后的阈值函数克服了软阈值函数中小波系数估计和原系数之间的固有偏差的缺点,也克服了硬阈值函数不连续的特点。实验结果表明,采用改进的阈值函数进行图像去噪处理后,图像具有更高的峰值信噪比,均方差更小,同时对软阈值函数的细节模糊现象和硬阈值函数的Gibbs现象都有不同程度的改良。
This paper introduces the traditional wavelet threshold denoising method of denoising prin -ciple.Wavelet threshold denoising is key threshold settings and threshold function selection .In this pa-per, the conventional soft threshold and hard threshold function in the conventional image denoising de -fects in is analyzed , and based on this , this paper proposes an improved threshold function , the improved threshold function to overcome the soft threshold function of small and medium -sized wave coefficient es-timation and the coefficient of the inherent deviation between faults , but also overcome the hard threshold with discontinuous function characteristics .The experimental results show that the improved threshold function image denoising processing , image has higher peak signal to noise ratio （ PSNR） , mean square error smaller , and the soft threshold function details fuzzy phenomena and hard threshold function of Gibbs phenomenon have different degrees of improvement .
出处
《工业仪表与自动化装置》
2013年第5期80-83,共4页
Industrial Instrumentation & Automation
关键词
小波阈值去噪
阈值函数
小波系数
峰值信噪比
wavelet threshold denoising
threshold function
wavelet coefficient
PNSR
