The methods of image analysis for allocation of road marking and estimation of its wear are considered. Based on the results of the analysis of markup technologies, an approach based on the use of a CRF graph is proposed with the subsequent removal of areas not related to markup and the union of areas belonging to the same markup line. Processing video of the road involves the following stages: allocation in the frame of the roadway analysis area, recognition of road marking in the selected zone, calculation of the lengths of marking sites, de-termination of the degree of wear of the markup, construction of a sheet of markup defects. To determine the degree of wear of the markup, a markup scheme in JSON format is used, which lists which types of markings should be applied on each section of the road. In the defective list, the coordinates of the road sections are kept, indicating the degree of wear of the markup and the number of video frames that correspond to these sec-tions. This list can be used for video viewing of sites where the marking lines are worn out, estimations of the necessary resources for their repair, fitting of the location of the existing markup to the road marking scheme and rules for its application.

Authors: V. V. Vihlyev, S. A. Kulakov, G. V. Razumovski, E. V. Tuomas

Direction: Informatics and Computer Technologies

Keywords: Road marking recognition, CRF graph, markup scheme, markup wear, bill of defects

View full article