Real-time detection of road markings for driving assistance applications

For a Driving Assistance System, which is designed to improve safety on the roads, knowledge about the type of lane border markings and other painted road objects is required. Furthermore, information about the position of the painted objects can be used by other systems to create a correct perception of the road. This paper describes a Lab VIEW based system for detection, measurement and classification of painted objects. The system is able to calibrate monocular video sources, define the features of the painted objects that we look for, and detect them in real time using geometric pattern matching and edge detection. The solution can be easily deployed on a dedicated real-time computing architecture, to be used as a standalone driving assistance sensor.