High-Speed Obstacle Detection for Automated Highway Applications Thesis Proposal

Highway obstacle detection is a challenging problem. Highways present an unknown and dynamic environment with real-time constraints. In addition, the high speeds of travel force a system to detect objects at long ranges. Although there are a number of methods that can successfully detect moving vehicles, the more difficult problem of finding small, static road debris such as tires or crates remains unsolved. Systems such as the Automated Highway System (AHS) which demand high levels of safety are not feasible unless these critical problems are addressed. Although the problem of detecting static obstacles has been tackled in both the crosscountry and indoor mobile robot navigation literature, these systems have operated at low speeds (5 10 mph or less) and short range. This thesis will improve on the current state-of-the-art, by demonstrating how small static road debris can be safely detected at long distances and high speeds. In particular, it will focus on using two sensor modalities: laser reflectance and stereo vision. Laser reflectance, to our knowledge, has not been used for obstacle detection before. The thesis will show that reliable detection can be achieved by using the right methods (sensitive enough) and the right models (no more complicated than necessary) for both road and sensor. We will develop two detection systems, laser and stereo-based, which can detect 20 cm high obstacles at 60 meters. The first system uses laser intensity to provide a more direct means of measuring surface orientation than traditional laser range-based processing: vertical obstacles should provide stronger laser returns than the horizontal road. The second system proposed is a predictive, model-based stereo method. Accurate modeling of the road and CCD sensor will enable obstacle detection without expensive 3-D reconstruction.