论文部分内容阅读
利用小波变换模极大值原理对信号去噪之后，如何由保留下来的模极大值点恢复出满意的重构信号，是一个重要课题。本文首先分析模极大值与小波系数之间的内在关系，提出了模极大值实际上是小波系数在特定意义下的离散采样；然后给出了一种对模极大值进行预处理的方法，由此得到了一组新的伪模极大值序列；利用这组伪模极大值序列，提出了一种新的重构小波系数的分段三次样条播值（ＰＣＳＩ）新算法，该算法程序简单，易实现，克服了交替投影（ＡＰ）法计算量大、程序复杂等缺点；最后给出一个应用实例，实验结果表明，与经典的交替投影法相比，本文提出的ＰＣＳＩ算法可获得更高的重构信号信噪比增益和更小的相对均方误差，它是一种实际、有效的算法。
After denoising the signal by the principle of wavelet transform modulus maxima, how to recover a satisfactory reconstructed signal from the retained modulus maxima is an important issue. In this paper, we first analyze the intrinsic relationship between modulus maxima and wavelet coefficients and propose that modulus maxima are actually discrete samples of wavelet coefficients in a certain sense. Then, we give a method to preprocess the modulus maxima A new set of pseudo-maxima sequences are obtained. Based on the set of pseudo-maxima sequences, a new PCSI algorithm for reconstructing wavelet coefficients is proposed The algorithm is simple and easy to implement, and overcomes the shortcomings of the alternating projection (AP) method such as large computational complexity and complex program. Finally, an application example is given. The experimental results show that compared with the classical alternative projection method, the proposed PCSI algorithm It can obtain a higher signal-to-noise gain of reconstructed signal and a smaller relative mean square error, which is a practical and effective algorithm.