PROJECT TITLE :

Real-Time Global Localization of Robotic Cars in Lane Level via Lane Marking Detection and Shape Registration

ABSTRACT:

During this paper, we tend to propose an correct and real-time positioning method for robotic cars in urban environments. The proposed methodology uses a strong lane marking detection algorithm, and an efficient form registration algorithm between the detected lane markings and a GPS-based mostly road form prior, to boost the robustness and accuracy of the global localization of a robotic automotive. We tend to show that, by formulating the positioning downside during a relative sense, we tend to will estimate the worldwide localization of a automotive in real time and certain its absolute error in the centimeter level by a cross-validation theme. The cross-validation theme integrates the vision-based mostly lane marking detection with the shape registration, and it improves the accuracy and robustness of the general localization system. The GPS localization will be refined by using lane marking detection when the GPS suffers from frequent satellite signal masking or blockage, whereas lane marking detection is validated and completed by the GPS-primarily based road shape prior when it does not work well in adverse weather conditions or with poor lane signatures. We extensively evaluate the proposed method with one forward-wanting camera mounted on an autonomous vehicle that travels at sixty km/h through many urban street scenes.