基于重排Gabor变换的高分辨率谱分解

High-resolution spectrum decomposition based on reassigned Gabor transform

地震信号是随机信号,频率成分随着时间变化而变化。Gabor变换是加窗傅里叶变换的一种,被广泛用于信号的时频分析,但分辨率有限。本文引入一种基于能量重排的Gabor变换,在时频平面上对能量关于其重心进行重排,提高了时频分析的分辨率。通过理论算例和实际数据证明,重排后的Gabor变换谱分解在烃类检测中有更好的时频分辨率,分层更准确,更精细地指示了烃类的存在。

The seismic signal is random signal. Its frequency is variable with time. Gabor transform method, a STFT being used in time-frequency analysis, is limited in time-frequency resolution. We introduce a reassigned Gabor transform, which assigns the total energy to the center of gravity on the time-frequency plane, to improve the resolution. The theoretical and real data examples show that the reassigned Gabor method has much higher time-frequency resolution, providing more detailed stratification and more accurate detection of the hydrocarbons.