摘要
构造一种新的小波半软阈值函数,在阈值范围内对小波系数进行细化处理,更符合自然信号的复杂连续特性;同时引入混沌搜索方法,对阈值可调参数进行遍历搜索,在硬阈值和软阈值之间得到小波系数的最佳估计。采用均方误差(MSE)和信噪比(SNR)对去噪效果进行评价,仿真结果表明该方法保留的小波系数能够恢复出更多的原始信号细节,改进了滤波效果,提高了去噪质量。
This paper constructs a new wavelet half-soft threshold function to finish refining treatment of the wavelet coefficients within the threshold, which is more suitable to the complex and continuous characteris- tics of the natural signal. Meanwhile, the chaotic searching method is introduced and the adjustable parame- ters of the threshold function is searched entirely. The study obtains the best estimate of the wavelet coeffi- cients between the hard and soft threshold. Using mean square error （MSE） and signal to noise ratio （SNR） to evaluate the de-noising effect, simulation results show that more details of the original signal can be restored by the retained wavelet coefficients obtained by method in this paper, and filtering effects as well as de-noising quality can be improved.
出处
《沈阳航空航天大学学报》
2013年第1期57-60,共4页
Journal of Shenyang Aerospace University
基金
国家自然科学基金(项目编号:60804025)
国家自然科学基金(项目编号:61074090)
航空基金(项目编号:2011ZD54011)
关键词
小波变换
阈值去噪
混沌
遍历性
wavelet transform
threshold de-noising
chaos
ergodicity
