基于低信噪比条件下新型 Seislet 变换的阈值去噪方法

摘要

Seislet transform is a wavelet-like transform based on wavelet base that analyzes seismic data along variable local event slopes.The calculation of the local event slopes is the key step of this method.There are many kinds of methods to calculate the local slopes.But in the condition of low SNR, these methods have some limits.We propose a method that is suitable for low SNR based on the definition of the slopes of t-x relationship in CMP traces.Comparing the slopes of t-x relationship in CMP traces with PWD method,we can see that the new method is more accurate in calculating the local slopes.Apply these slopes to Seislet frame to establish a new Seislet transform that characterize low SNR data.In seismic data processing,we introduce one method which works well for acoustic signals. We propose the new threshold method with Seislet transform.It is not only suitable for seismic signals, but also better than the traditional threshold method to improve SNR.The results show that the combination of the new Seislet transform and improved threshold method can extract signal in the condition of low SNR effectively.%Seislet 变换是一种类小波变换方法，主要根据小波基沿地震同相轴的局部倾角方向来分析数据，其中，局部地震倾角的表征是该方法的核心。局部倾角的求取方法有很多种，但是往往在低信噪比的条件下存在着一些局限。根据共中心点道集中基于时距关系的地震倾角定义，提出一种适应于低信噪比条件下的倾角求取方法。对比基于时距关系与平面波分解滤波器计算出的局部地震倾角，结果证明，该方法能更加准确地表征低信噪比条件下同相轴的倾角信息。将基于时距关系的局部地震倾角用于 Seislet 框架，建立表征低信噪比数据的新型 Seislet 变换方法。在地震数据处理中，引入语音信号中改进的阈值方法。结合新型 Seislet 变换，提出阈值去噪方法，此方法不但适于地震数据，而且在提高信噪比方面也优于传统的阈值去噪方法。实际数据处理的结果验证了新型 Seislet 变换与改进阈值去噪方法的组合能够有效地解决低信噪比条件下的信号提取任务。