A fractal-based algorithm for detecting first arrivals on seismic traces
A fractal-based algorithm for detecting first arrivals on seismic traces
复制标题
DOI：
10.1190/1.1444030
复制
发表时间：
1996-07-01
期刊：
影响因子：
3.3
通讯作者：
List, RD
中科院分区：
文献类型：
--
作者：
Boschetti, F;Dentith, MD;List, RD
提出了一种用于自动拾取地震初至波的新算法，该算法通过分析沿地震道的分形维数变化来检测信号的存在。发现“分割法”是计算分形维数最合适的方法。发现在从噪声到信号加噪声的转变（即初至波）附近，维数会发生变化。这种变化的性质因道而异，但总是能发现可检测到的变化。该算法已在具有不同信噪比的真实数据集上进行了测试，并将结果与使用先前发表的算法所获得的结果进行了比较。通过对其参数进行适当调整，基于分形的算法被证明比所有其他算法都更准确，尤其是在存在显著噪声的情况下。分形方法被证明能够容忍高达平均信号幅度80%的噪声。然而，基于分形的算法比其他方法慢得多，因此仅适用于低信噪比的数据集。
A new algorithm is proposed for the automatic picking of seismic first arrivals that detects the presence of a signal by analyzing the variation in fractal dimension along the trace. The ''divider-method'' is found to be the most suitable method for calculating the fractal dimension. A change in dimension is found to occur close to the transition from noise to signal plus noise, that is the first arrival. The nature of this change varies from trace to trace, but a detectable change is always found to occur. The algorithm has been tested on real data sets with varying SIN ratios and the results compared to those obtained using previously published algorithms. With an appropriate tuning of its parameters, the fractal-based algorithm proved more accurate than all these other algorithms, especially in the presence of significant noise. The fractal method proved able to tolerate noise up to 80% of the average signal amplitude. However: the fractal-based algorithm is considerably slower than the other methods and hence is intended for use only on data sets with low SIN ratios.