Ego-Lane Estimation by Modeling Lanes and Sensor Failures

Main Project Page - The Road Layout Estimation Project

Abstract - In this paper we present a probabilistic lane-localization algorithm for highway-like scenarios designed to increase the accuracy of the vehicle localization estimate. The contribution relies on a Hidden Markov Model (HMM) with a transient failure model. The idea behind the proposed approach is to exploit the availability of OpenStreetMap road properties in order to reduce the localization uncertainties that would result from relying only on a noisy line detector, by leveraging consecutive, possibly incomplete, observations. The algorithm effectiveness is proven by employing a line detection algorithm and showing we could achieve a much more usable, i.e., stable and reliable, lane-localization over more than 100Km of highway scenarios, recorded both in Italy and Spain. Moreover, as we could not find a suitable dataset for a quantitative comparison of our results with other approaches, we collected datasets and manually annotated the Ground Truth about the vehicle ego-lane. Such datasets are made publicly available for usage from the scientific community.

Published in the proceedings of the IEEE Intelligent Transport System Conference, ITSC 2017.

Download the presentation slides here: Ego-Lane Estimation by Modeling Lanes and Sensor Failures (Slides in PDF format) (23 downloads)

The first datasets and the associated ground truth data will be available soon; images will be in the standard PNG file format, the ground truth in a text file. A ROS bag file will be available on request. The images were recorded using a PhotonFocus MV1-D1312-40-GB-12 camera at the 1312x540 resolution on the A4 highway (from Milano to Bergamo and vice versa), Italy, in real traffic conditions, by IRAlab [link]. Before links are made available, send email to domenico .dot. sorrenti @at@ unimib .dot. it to ask for the material.

The second dataset was recorded in Spain, on the A2 highway nearby Alcalá de Henares, by the INVETT Research Group [link].

an image from the A4 dataset
An example image from the A4 (Milano - Bergamo) dataset


an image from the A2
An example image from the A2 (Alcalà de Henares - Madrid) dataset