An algorithm for detecting roads and obstacles in radar images

This paper describes an algorithm fur detecting roads and obstacles in rada r data taken from a millimeter-wave imaging platform mounted on a stationar y automobile. Such an algorithm is useful in a system that provides all-wea ther driving assistance. Road boundaries are detected first, The prior shap e of the road boundaries is modeled as a deformable template that describes the road edges in terms of its curvature, orientation, and offset. This te mplate is matched to the underlying gradient field of the radar data using a new criterion, The Metropolis algorithm is used to deform the template so that it "best" matches the underlying gradient field. Obstacles are detect ed next. The radar returns from image pixels that are identified as being p art of the road are processed again, and their power levels are compared to a threshold. Pixels belonging to the road that return a significant (great er than a fixed threshold) amount of incident radar power are identified as potential obstacles. The performance of the algorithm on a large all-weath er data set is documented. The road edges and obstacles detected are consis tently close to ground truth over the entire data set. A new method for com puting the gradient field of radar data is also reported, along with an exp osition of the millimeter-wave radar imaging process from a signal-processi ng perspective.