Volume XL-3/W2
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XL-3/W2, 227-233, 2015
https://doi.org/10.5194/isprsarchives-XL-3-W2-227-2015
© Author(s) 2015. This work is distributed under
the Creative Commons Attribution 3.0 License.
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XL-3/W2, 227-233, 2015
https://doi.org/10.5194/isprsarchives-XL-3-W2-227-2015
© Author(s) 2015. This work is distributed under
the Creative Commons Attribution 3.0 License.

  10 Mar 2015

10 Mar 2015

ON OUTLIER DETECTION IN A PHOTOGRAMMETRIC MOBILE MAPPING DATASET

C. Taglioretti1, A. M. Manzino1, T. Bellone1, and I. Colomina2 C. Taglioretti et al.
  • 1DIATI Department, Politecnico di Torino, Corso Duca degli Abruzzi 24, 10129 Torino, Italy
  • 2Geomatics Division, GEON Department, Centre Tecnològic Telecomunicacions Catalunya, Parc Mediterrani de la Tecnologia, (PMT) Building B4, Av. Carl Friedrich Gauss 7, 08860 Castelldefels, Spain

Keywords: Mobile Mapping, relative orientation, outlier detection, robust statistical, RANSAC, Forward Search ABSTRACT:

Abstract. Various types of technology are used for Terrestrial Mobile Mapping (TMM) such as IMU, cameras, odometers, laser scanner etc., which are integrated in order to determine the attitude and the position of the vehicle in use, especially in the absence of GNSS signal i.e. in an urban canyon.

The aim of this study is to use only photogrammetric measurements obtained with a low cost camera (with a reduced focal length and small frames) located on the vehicle, in order to improve the quality of TMM solution in the absence of a GNSS signal. It is essential to have good quality frames in order to solve this problem. In fact it is generally quite easy to extract a large number of common points between the frames (the so-called ‘tie points’), but this does not necessarily imply the goodness of the matching quality, which might be uncorrected due to the presence of obstacles that may occlude the camera sight. The Authors used two different methods for solving the problem of the presence of outliers: RANSAC and the Forward Search.

In this article the Authors show the results obtainable with good quality frames (frames without occlusions) and under difficult conditions that simulate better reality.