Detection of dynamic obstacles out of the line of sight for autonomous vehicles to increase safety based on shadows

Abstract

Moving obstacles occluded by corners are a potential source for collisions in mobile robotics applications such as autonomous vehicles. In this work, we address the problem of anticipating such potential collisions by proposing a vision-based detection algorithm for obstacles which are outside of a vehicle's direct line of sight. Our method detects shadows of obstacles hidden around corners and automatically classifies these unseen obstacles as "dynamic" or "static". We evaluate our proposed detection algorithm on real-world corners and a large variety of simulated environments to assess generalizability in different challenging surface and lighting conditions. For the image registration step we compare a visual odometry method (i.e. DSO) with a fiducial marker system (i.e. AprilTags). The mean classification accuracy on simulated data is around 80% and on real-world corners approximately for both image registration methods 70%. Additionally, we integrate our detection system on a full-scale autonomous wheelchair and demonstrate its feasibility as an additional safety mechanism through real-world experiments. We release our real-time-capable implementation of the proposed ShadowCam algorithm and the dataset containing simulated and real-world data under an open-source license.