Convolutional Recurrent Network for Road Boundary Extraction
This work is significant for self-driving cars, providing precise road boundary information for safe navigation.
This paper addresses the problem of drivable road boundary extraction from LiDAR and camera imagery for high-definition maps. The proposed method achieves perfect topology of road boundaries 99.3% of the time with high precision and recall in a large North American city.
Creating high definition maps that contain precise information of static elements of the scene is of utmost importance for enabling self driving cars to drive safely. In this paper, we tackle the problem of drivable road boundary extraction from LiDAR and camera imagery. Towards this goal, we design a structured model where a fully convolutional network obtains deep features encoding the location and direction of road boundaries and then, a convolutional recurrent network outputs a polyline representation for each one of them. Importantly, our method is fully automatic and does not require a user in the loop. We showcase the effectiveness of our method on a large North American city where we obtain perfect topology of road boundaries 99.3% of the time at a high precision and recall.