Robust Multiple Car Tracking with Occlusion Reasoning

Abstract

We address the problem of occlusion in tracking multiple 3D objects in a known environment. For that purpose we employ a contour tracker based on intensity and motion boundaries. The motion of a contour enclosing the image of a vehicle is assumed to be well describable by an affine motion model with a translation and a change in scale. Contours are represented by closed cubic splines the position and motion of which are estimated along the image sequence. In order to employ linear Kalman Filters we decompose the estimation process in two filters: one for estimating the affine motion parameters and one for estimating the shape of the contours of the vehicles. Occlusion detection is performed by intersecting the depth ordered regions associated to the objects. The intersection part is then excluded in the motion and shape estimation. Occlusion reasoning also improves the shape estimation in case of adjacent objects where shape estimates can be corrupted by image data of other objects. In this way we obtain robust motion estimates and trajectories for vehicles even in the case of occlusions, as we show in some experiments with real world traffic scenes.