Data fusion algorithms for lane departure warning systems

Abstract

This paper discusses data-fusion algorithms for computing the “Time-to-Lane-Crossing” (TLC) of a vehicle traveling along a lane on the basis of road images, collected by an on-board video-camera, and kinematic data coming from car sensors. Such algorithms are instrumental to built Lane Departure Warning Systems (LDWS) with enhanced predictive capabilities and tolerant to a certain amount of errors and outliers in video-camera and sensors data flows. Two Kalman filters are outlined and used for data fusion purposes. Simulations within the Carsim <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">®</sup> simulation tool have been undertaken to verify the effectiveness of these filters during critical driving scenarios.