ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Publications Copernicus
Articles | Volume II-5/W2
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., II-5/W2, 157–162, 2013
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., II-5/W2, 157–162, 2013

  16 Oct 2013

16 Oct 2013

A positioning free calibration method for mobile laser scanning applications

R. Le Scouarnec1, T. Touzé2, J. B. Lacambre1, and N. Seube2 R. Le Scouarnec et al.
  • 2iXBlue, Marly-le-roi, France
  • 1ENSTA Bretagne, Dept STIC/OSM, Brest, France

Keywords: Mobile laser scanning, Calibration, Boresight, Positioning-free, Plane features

Abstract. Mobile laser scanning are likely to find more and more applications for high density 3D environmental data. A mobile laser scanning system is composed by three subsystems: a GNSS (Global Navigation Satellite System) that provides position information, an INS (Inertial Navigation System) for attitude determination, and a LiDAR (Light Detection And Ranging). The accuracy of the geolocated LiDAR returns depends on the accuracy of each instrument but also on the bore-sighting parameters and the lever arms between the instruments. Indeed, an imperfect calibration may lead to systematic errors. Calibration may then become the limiting factor of Terrestrial Laser scanning if it is not tackled seriously. Moreover [Ø], it is important to have a reliable value of the calibration precision. This paper presents a new positioning free procedure for the estimation of the LiDAR bore-sighting parameters. Since this method is static, lever arms do not affect the boresight calibration and positioning is not required. That makes the methodology immune to GPS errors. Finally, since it is based on a rigorous mathematical model, it can provide a reliable boresight quality factor. First, the boresight determination problem is explained and existing calibration procedures are introduced. After having explained their drawbacks, a new procedure that tries to overcome these limitations is described. Tests from simulations and real datasets are also presented to illustrate our approach.