节点文献
基于数字图像处理的地震波初至时间检测
First Arrival Picking Based on Image Processing Methods
【摘要】 地震波初至时间的检测是地震信号处理中的一个重要环节。在地震波接收记录上,初至波振幅大于噪声振幅,初至时间位于纯噪声信号与地震有效信号和噪声之叠加信号之间的分界处。根据这些特征,本文把地震信号幅值数据处理成256级灰度图,再把灰度图二值化,然后通过检测二值图像的边缘,获得地震波的初至时刻。对于存在较多的连续异常记录道,或其它强干扰信号的地震波记录,提出了二次或多次用不同阈值对图像进行二值化的检测方法。对实际地震记录的处理表明,本文方法具有较高的精度和效率。
【Abstract】 This paper discusses a new method of first arrival picking. The method is based on image processing methods.It is proposed for the purpose of automatic and accurate picking of large volumes of first arrival picks for refraction static analysis.The first step converts the seismic data into a grey image.The second step converts the grey scale image to a binary image.At last simple secondary pick logic can be employed to calculate a final pick time.In the course of picking,the paqer discusses a new method to influence the accuracy of picking first arrival,converts the grey scale image to more than 2 binary images by different threshold value.The experimentation has proven that the new method has great speed and efficiency.
【Key words】 first arrivals on seismic data; automatic picking; binary image; edge extraction;
- 【文献出处】 计算机与现代化 ,Computer and Modernization , 编辑部邮箱 ,2007年04期
- 【分类号】TP391.41
- 【被引频次】9
- 【下载频次】201