二维非平稳信号的小波谱均衡分析方法

小波变换对信号的分时分频的精细表达和多分辨率的全面把握，使我们能更深入地认识信号和噪声的特性，便于基于小波域进行更有效的高分辩率的处理工作；而谱均衡处理技术是提高信号纵向分辨率的有效方法．本文结合上述两种处理手段，提出了一种小波谱均衡方法：先将二维非平稳信号进行小波分频处理，得到不同尺度的信号；然后将各尺度上的信号分别进行谱均衡处理，再将经谱均衡处理的不同尺度信号进行重建，从而得到高分辨率信号．利用小波谱均衡进行实际信号处理，处理结果表明该方法的处理效果要好于常规的谱均衡方法，它既能提高二维非平稳信号的纵向分辨率，同时又能保持原信号的固有特征．

Spectrum equalization based on wavelet transform for 2D non-stationary signal

Due to the refined expression and multiresolution analysis in wavelet transform domain for a signal, the characteristics of the signal and noise can be recognized. So it is easy for us to perform efficient high resolution processing. The processes of spectrum equalization can also improve the vertical resolution of seismic data. Each of the two methods has advantages and shortcomings. This paper combines the two methods into a new one named spectrum equalization based on wavelet transform. First, the signals with different scales are obtained when non-stationary signal are processed by wavelet transform. Then these signals with different scales are processed separately by spectrum equalization method. Finally, scaled signals that have been equalized are reconstructed to form a new signal with high resolution. Using the above new method to process the given signals has led to the result that new method has finer effects than an ordinary spectrum equalization method and it can both improve the signal's vertical resolution and keep the intrinsic characteristics of the original signal.