Curb Detection Based on Elevation Maps from Dense Stereo

A new approach for the detection of curbs (borders of relevant traffic isles, sidewalks, etc) is presented, in the context of urban driving assistance systems. The 3D data from dense stereo is transformed into a rectangular elevation map. The 3D resolution and uncertainty of the stereo sensor are evaluated and used in the detection process. Edge detection is applied on the elevation map in order to highlight height variations. Relevant lines are extracted using a Hough accumulator. Each relevant line is validated as a curb by applying a set of proposed conditions. The algorithm was evaluated in an urban scenario. It works in real time and provides robust detection of curbs.