Anytime Lane-Level Intersection Estimation Based on Trajectories of Other Traffic Participants

Abstract

Estimating and understanding the current scene is an inevitable capability of automated vehicles. Usually, maps are used as prior for interpreting sensor measurements in order to drive safely and comfortably. Only few approaches take into account that maps might be outdated and lead to wrong assumptions on the environment. This work estimates a lanelevel intersection topology without any map prior by observing the trajectories of other traffic participants.We are able to deliver both a coarse lane-level topology as well as the lane course inside and outside of the intersection using Markov chain Monte Carlo sampling. The model is neither limited to a number of lanes or arms nor to the topology of the intersection.We present our results on an evaluation set of 1000 simulated intersections and achieve 99.9% accuracy on the topology estimation that takes only 36ms, when utilizing tracked object detections. The precise lane course on these intersections is estimated with an error of 15cm on average after 140ms. Our approach shows a similar level of precision on 14 realworld intersections with 18cm average deviation on simple intersections and 27cm for more complex scenarios. Here the estimation takes only 113ms in total.

Lane Graph Estimation for Scene Understanding in Urban Driving

“…Similarly, Bastani et al [2] propose to postprocess the output of a binary segmentation model through morphological thinning and the Douglas-Peucker method. Based on simulated data, Meyer et al [16] propose a lanelevel intersection estimator, leveraging birds-eve-view tracking of traffic participants for intersection graph estimation. Li et al [11] use an RNN to build building polygons and road graphs from overhead images, while Chu et al [6] propose to iteratively build a road graph from overhead images.…”

Automatic Construction of Lane-level HD Maps for Urban Scenes

“…For example, in [18] [19], the authors used vehicle trajectories acquired from on-board sensor such as stereo camera or LIDAR. Later, Meyer et al used simulated vehicle trajectories and employed a Markov chain Monte Carlo sampling to reconstruct the lane topology and geometry at both real and simulated intersections [20], [21]. In [22], [23], the authors used collected GPS data loaded on fleet vehicles.…”

Semantic Segmentation of Urban Scenes Using Spatial Contexts

“…Unlike the indoor environments, the outdoor environment is less affected by a Manhattan world assumption [20], research for modeling the outdoor layout has been done [21]- [25]. Jo and Sunwoo [21] and Alvarez et al [22] modeled the parameterized roadmap for autonomous car driving using the linear interpolation of points from the trajectory of the road or B-spline curve fitting.…”

“…Geiger et al [23] proposed a probabilistic model for multi-object traffic scene understanding of video input sequences that jointly infers the urban scene layout as well as the position and orientation of moving objects using visual evidence. Meyer et al [25] suggested the generalized version of road layouts of Geiger's work run in real-time.…”