Evaluation of Automatic Road Extraction

Internal self-diagnosis and external evaluation of the obtained results are essential for any automatic system. In the long run these factors are of major importance for the relevance of the system for practical applications. Obviously, this statement is also true for image analysis in photogrammetry and remote sensing. However, so far only relatively little work has been carried out in this area. This paper deals with the external evaluation of automatic road extraction algorithms by means of comparison to manually plotted linear road axis used as reference data. The comparison is performed in two steps: (1) Matching of the extracted primitives to the reference network; (2) Calculation of quality measures. Each part 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, the matching process is an important part of the evaluation process. The quality measures proposed for the automatically...