Automatic velocity estimation in GPR data and migration

Ground Penetrating Radar is a widely used technique for detecting objects in the subsurface. However, the objects typically appear as hyperbolic curves in the raw data's time domain. Therefore, the size and the location of the objects are hard to interpret from the raw data. Migration is a helpful technique to determine the subsurface velocity and to transform the raw data into an image where the object's size and location are easier to interpret. In this thesis, we discuss a strategy for automatic migration. We apply the automatic strategy with a few standard migration methods, Kirchhoff, Phase-shift, and f-k Stretch migration. The automatic strategy was tested on both real and synthetic data. The results showed that the automatic migration strategy successfully could improve the interpretation of both datasets. In the synthetic data test, we could estimate the location of objects with a relative error of about 2-12%