Lidar-Assisted Georeferencing of Multi-Channel GPR Data

Multi-channel GPRs can potentially vastly increase coverage and frequency of road condition monitoring, but only if there is an accurate way to georeference its data so that multiple passes can be accurately aligned. Lidar data, when collected together with IMU and GNSS, can reveal errors in the IMU and GNSS trajectory. By adjusting the trajectory such that these errors are reduced, we effectively obtain a more accurate trajectory, which can also be used to more accurately align multichannel GPR data. This paper demonstrates the benefits of using such a Lidar-assisted georeferencing approach. Not only does such an approach reduce distortion when stitching C -scans from separate drive passes, it can also precisely align same-location different-pass B-scans so that we can disambiguate actual change from positioning error.