基于互相关函数对钻孔雷达层析成像的改进

Improved Borehole-GPR Tomography Based on Cross-Correlation

常用的旅行时提取方法在针对大收发角度数据进行提取时,因其低信噪比而显得较为艰难且不够准确。提出一种新的基于互相关函数的方法对旅行时提取方法进行改进。首先对数据进行归一化并按角度对数据分类;然后对相同收发角度数据与该角度信噪比最大的数据进行互相关处理并叠加,从而得到该收发角度数据的参考波形;最后根据参考波形的初至波到达时间对该角度数据进行互相关处理,得到每道数据的初至波旅行时。通过计算机模拟的结果和对RAMAC雷达采集的实际数据进行分析处理的结果,对该方法的准确性进行了论证,与常规旅行时提取法得出的层析成像图相对比,该方法得到的层析成像图更为精确、清晰,能为判断地下异常提供更为有力的帮助。

First-arrival times at high transmitter-receive angles are often very difficult and inaccurate to pick up by conventional methods because of low signal/noise in the data.We introduce a new traveltime picking scheme based on cross-correlation.First,we normalize each trace by signal maximum and sort the data into common-ray-angle gathers.Then,we align traces in each common-ray-angle gather by cross correlating with the other trace of the highest signal-to-noise ratio,determine mean trace for each angle gather.Finally,we pick first-arrival time on each mean trace,cross-correlate each trace in data set with appropriate reference waveform for that angle,and determine first-arrival time.In the test of synthetic data and field data,the approach improves the results.Compared with velocity tomography derived from common time picking technology,the new velocity tomography is more accurate,more clearly,and more helpful to distinguish the target under the ground.