基于图像分析的双台面波相速度频散曲线快速提取方法

A quick tracing method based on image analysis technique for the determination of dual stations phase velocities dispersion curve of surface wave

双台波形的互相关方法是面波相速度频散曲线提取的一种常用方法。传统的方法精度不高，效率较低。本文提出了一种基于图像分析的频散曲线快速提取方法，采用图像显示互相关振幅矩阵，且提出了快速追踪整条频散曲线的算法，并用 Matlab 编写了交互式的处理软件。基于这种方法编写的资料处理软件提高了双台相速度频散曲线提取的速度和精度，为大批量处理地震数据资料提供了坚实的软件基础。基于图像分析的方法可以更为清晰地展现出双台间的面波频散曲线的特征，对频散曲线的识别和分析也提供了更为有效的研究工具。

The cross-correlation of the dual stations waveform is a routine method for determining phase velocities dispersion curve of surface wave. Traditionally, the dispersion curve is delineated by hand on the contour map of cross-correlation amplitude matrix. Its precision and efficiency is very low. In this article, we propose a quick tracing method based on image analysis technique for the determination of the dispersion curve. We wrote interactive GUI software with Matlab. We adopt that the image displays the cross-correlation amplitude matrix. Base on it, we can be easy to identify the dispersion curve correctly. The technique provides an algorithm of the quick tracing entire curve and may pick up the entire dispersion curve quickly and precisely on the image of the cross-correlation amplitude matrix, at same time, the dispersion datum are saved in the file automatically. The processing software advances the precise and efficient for the picking up the dispersion curve of the dual stations, and provides a solid basis for the great deal processing the seismic datum. The image analysis method may clearly display the character of surface wave dispersion for the dual stations and provides an efficient approach to identify and analysis dispersion curve.