A threshold and correlation method based on wavelet transform is proposed for incoming noise originally from the non-stationary and multi-spike acoustic signal which affects subsequent analysis step. To optimize and improve the denoising, the algorithm could suppress the noise in high frequency and preserve the signal to the utmost extent by improving the spatial correlation algorithm and the soft threshold denoising algorithm, which introduces the correlation coefficient of between the scales into the soft threshold function by making use of the wavelet-coefficient difference of between signal and noise. And the proposed method could improve the noise judging mechanism, reduce the computational complexity and raise the denoising performance. The simulation results indicate that compared with the traditional correlation denoising method, the proposed method could steadily raise the signal-to-noise ratio and reduce the mean square error. improved signal-to-noise ratio and reduced the mean square error comparing with conventional correlation denoising method.%水声信号具有非平稳、多尖峰的特性,引入噪声会影响后续处理分析环节.为优化去噪算法,改善去噪效果,在空域相关算法和软阈值去噪算法的基础上进行改进,提出了一种基于小波变换的阈值相关去噪算法.利用信号与噪声在多尺度分解后的小波系数相关性不同的特点,将尺度间相关系数引入软阈值函数中,能抑制高频小波系数噪声部分的同时,最大程度保留信号边缘信息.该算法优化了相关算法的噪声判别机制,减少了运算复杂性,并改善了去噪效果.对仿真信号和实际信号去噪的结果显示,与常规相关去噪法相比,该方法能稳定提升信噪比,并降低均方误差.
展开▼
