Empirical Evaluation of Automatically Extracted Road Axes.pdf

EMPIRICAL EVALUATION OF AUTOMATICALLY EXTRACTED ROAD AXES Christian Wiedemann, Christian Heipke, Helmut Mayer Chair for Photogrammetry and Remote Sensing Technische Universit¨at M¨unchen, D–80290 Munich, Germany Tel: +49-89-2892-2676, Fax: +49-89-2809573 E-mail: fwiedgfchrisgfhelmutg@photo.verm.tu-muenchen.de Olivier Jamet I.G.N., Laboratoire MATIS 2, avenue Pasteur, 94160 Saint Mand´e, France E-mail: Olivier.Jamet@ign.fr Abstract Internal self-diagnosis and external evaluation of the obtained results are of major importance for the relevance of any automatic system for practical appl ications. Obviously, this statement is also true for automatic image analysis in photogrammetry and remote sensing. However, so far only relatively little work has been carried out in this area. This is mostly due to the moderate results achieved. Only recently automatic systems reached a state in which a systematic evaluation of the results seems to be meaningful. This paper deals with the external evaluation of automatic road extraction algorithms by com- parison to manually plotted linear road axes used as referen ce data. The comparison is performed in two steps: (1) Matching of the extracted primitives to the reference network; (2) Calculation of quality measures. Each step depends on the other: the less tolerant is matching, the less exhaustive the extraction is considered to be, but the more accurate it looks. Therefore, matching is an impor- tant part of the evaluation process. The quality measures proposed for the automatically extracted road data comprise completeness, correctness, quality, redundancy, p