摘要
在分析小波阈值降噪原理的基础上,为了克服硬阈值函数不连续和软阈值函数中估计小波系数与分解小波系数之间存在恒定偏差的问题,构造了一种改进的双曲线阈值函数.该方法通过调整参数可实现阈值函数形状的改变,采用对非重要小波系数的处理,不是均设为0,系数由多项式调节以接近理想小波系数,使得该阈值函数既具有软阈值函数连续、高阶可导的优点,又便于进行各种数学处理.经实验验证,这种方法不仅可以有效地去除噪声,而且还可以保留图像的细节信息,比传统阈值方法更接近最佳情况.
Based on analyzing the principle of wavelet threshold de-nosing,a hyperbola thresholding function is presented in order to overcome the discontinuance of the hard thresholding function and the constant deviation between the estimated wavelet coefficients and the decomposition wavelet coefficients in the soft thresholding function.The shape of new thresholding function can be changed visually by adjusting the parameters of hyperbola.The algorithm is improved by a non-important wavelet coefficient which is not supposed as zero but adjusted to approach its ideal wavelet coefficient by polynomial.The new thresholding function is simple in expression as continuous as the soft thresholding function and has a high order derivative which makes some kinds of mathematical disposals conveniently.At the same time,the new threshold function is more elastic than the soft-threshold and the hard-threshold functions.The simulation results show that the method can remove noise effectively,retain detail information of images,and can be closer than the traditional threshold method to the best situation.
出处
《中北大学学报（自然科学版）》
CAS
北大核心
2010年第6期625-630,共6页
Journal of North University of China(Natural Science Edition)
基金
山西高校科技研究开发项目(200713013)
关键词
小波阈值
双曲线阈值函数
峰值信噪比
均方误差
wavelet threshold
hyperbola thresholding function
power signal to noise ratio
mean square error
