CONSTRAINING GEOREFERENCED CONTROL POINTS BY OBJECT TO BE TRANSFORMED WITH APPLICATION TO TERRESTRIAL LASER SCANNING

With recent availability of terrestrial laser scanners to academia, applications have been made to investigating geologic questions which may have been infeasible by prior methods. New capability derives primarily from scale of number of measurements made (LiDAR point cloud data), allowing for rigorously constructed modeling (equations) as well as tracking subtle trends and features. Tight precision of scanner measurement relative to instrument accompanies increased challenge of integrating data between instruments or multiple sessions of same instrument: precision becomes an artifact if accuracy degrades across weaker links (georeferenced positioning) or less constrained composition (e.g. integrating two scan positions). In large scale composition of e.g. terrain data, some techniques (point cloud matching) available to counter misalignment artifacts amongst overlapping scan positions may introduce serially accumulated bias in the georeferenced positions of otherwise unbiased, independent scans. It is desirable to compromise as little as possible, thus the original georeferenced position and orientation (through its control) of the scanner should be obtained with the highest fidelity.

An investigation is proposed to a subtlety in whether the precision of the scanner itself can lend towards greater accuracy in its obtained georeferenced (GNSS) position. Free software utility and 3D transformation technique described by L.S. White at UTD perform georeferencing of scanner measurements through least-squares matching of (scanner + GPS) control points. Is preserved accuracy contingent on GPS accuracy of control points? Investigation is whether GPS accuracy of control points may be improved by constraining their relative positions via laser measurement. For emphasis, low accuracy (autonomous) GPS positions will feed the algorithm; obtained corrections compared with differential-GPS (kinematic) “truth.” Propose that scientific interest of this network adjustment is in question (non-obvious answer) of whether adjustment will improve ultimate accuracy of scanner georeferencing, since least squares derivation of (e.g. 6-D) transformation parameters is itself analogous to network adjustment by constraint via scanner measurement. Answer to be obtained in repeated survey of monument site.