End-to-End Deep Structured Models for Drawing Crosswalks
This work provides a highly automated solution for drawing crosswalk boundaries, which is significant for autonomous driving and mapping applications.
This paper addresses the problem of detecting crosswalks from LiDAR and camera imagery. By projecting inputs onto the ground surface and using CNNs for semantic cues, combined with road centerlines, the method achieves 96.6% automation in drawing crosswalk boundaries in a large city area.
In this paper we address the problem of detecting crosswalks from LiDAR and camera imagery. Towards this goal, given multiple LiDAR sweeps and the corresponding imagery, we project both inputs onto the ground surface to produce a top down view of the scene. We then leverage convolutional neural networks to extract semantic cues about the location of the crosswalks. These are then used in combination with road centerlines from freely available maps (e.g., OpenStreetMaps) to solve a structured optimization problem which draws the final crosswalk boundaries. Our experiments over crosswalks in a large city area show that 96.6% automation can be achieved.