Development of a System for Fixing Road Markings in Real Time

The purpose of this article is to analyze methods for detecting objects in a video stream (on video footage) and to develop a driver assistance system that allows detecting and recognizing objects using these methods in real time. The article describes a number of problems that exist in existing systems of fixing road infrastructure; elaboration of the topic and an overview of previous achievements; as well as the development of software for the system of fixing road markings (including a description of the methodology of the experiment, the process of calibration of the camera and perspective transformation, the implementation of detection and tracking of the road lane, and a description of the results of the software module).