Plane-based self-calibration procedure for terrestrial laser scanners
Terrestrial laser scanners provide a three-dimensional sampled (i.e. point cloud) representation of the surfaces of objects resulting in a very large number of points. They have great potential to improve the measurement and representation of remote and widespread objects for applications such as engineering metrology, cultural heritage recording, forestry, and so on. However, laser scanner measurements contain systematic errors such as range finder offset, collimation axis error, etc. Proper modelling and estimation of the corresponding calibration parameters is critical in order to remove these errors prior to performing a measurement task such as deformation monitoring.
A rigorous, point-based self-calibration method developed by the Curtin laser scanner research group has already been shown to be effective. However, it is very labour-intensive since it requires manual measurement of a large number of signalised targets. In this paper, we propose an automated, plane-based self-calibration method that reduces the manual labour needed in the point-based method.
After outlining the principles and mathematical models of the proposed method, the subject of model identification is addressed. Tests with simulated datasets have revealed that the residual patterns from the plane-based method are markedly different from those of the point-based method. The ramification of this outcome is that systematic error identification, an important process for new instrumentation such as terrestrial laser scanners, is not straightforward. In addition, the precision and accuracy of the proposed method with terrestrial laser scanner datasets are presented, analysed and rigorously compared with the results from the point-based self-calibration method.