Currently, software applications are widely used, among which we mention the navigation systems on vehicles , traffic monitoring and management systems, etc., which require the use of road databases. These databases must contain accurate, updated information and populated with sufficient descriptive values. This article presents a semi-automatic algorithm for extracting roads from high-resolution aerial and satellite recordings, based on the weighted correlation of cross-sectional profiles. The algorithm uses as initial data two starting points with the help of which the road orientation and the template profile are obtained. Also, the operator must set the threshold value of the correlation coefficient between the transverse profiles, the pitch and the search angle, the length of the transverse profile and the maximum number of rejections Compared to other semi-automatic road vectorization algorithms, this algorithm is less sensitive to changes radiometrics from the ends of the profile, due to the assignment of higher weights to the central pixels. In this article is proposed an algorithm for semi-automatic vectoring of roads on color aerial images with submetric resolution, an algorithm that uses the weighted correlation of road cross profiles with a template profile. The operator must provide two points successive start, as close to the center of the road as possible, with the help of which the direction of start and template profile.

road, resolution, vectorization, analyze

geodesy engineering

Presentation: poster