【摘要】 在地震资料中,噪声干扰严重影响了有效信号的提取,为此必须进行信噪分离处理.本文提出一种基于Curvelet变换和KL变换相结合的软硬阈值折衷处理方法.首先对地震数据进行Curvelet变换,然后对各尺度系数选取适当阈值压制噪声干扰,再利用KL变换提取数据中的相干有效信号,最后重构得到去噪后的记录.经合成记录和实际地震资料处理实验证明,该方法与小波变换法相比较,更能有效进行信噪分离,提高地震剖面信噪比和分辨率.更多还原

【Abstract】 In seismic data, noises seriously affect the extraction of significant signals, so the denoising processing is necessary. A method of signal noise separation is proposed based on a combination of Curvelet transform and KL transform. This paper first uses the Curvelet method to decompose seismic data. The soft and hard thresholds on scales of the Curvelet domain are selected for attenuation of noises, and then we makes use of KL transform to extract correlative significant signals. Lastly we reconstruct the seismic data processed. The synthetic and real seismic data processing experiments show that the method in this paper can efficiently separate signal and noise, and improve seismic section signal-to-noise ratio and seismic resolution compared with the wavelet method.更多还原