Skip to Main Content
Perception of the surrounding environment is one of the many tasks an automated vehicle has to achieve in complex and ever-changing surroundings. This typically includes several distinct sub-tasks, such as map-building, localisation, static obstacles detection, pedestrian detection,... Some of these tasks are nowadays very well known, such as map-building, whereas the perception, localisation and classification of moving objects from a moving vehicle are in many aspects a work in progress. In this paper, we propose a vision-based approach built on the extensive tracking of numerous visual features over time from a stereo-vision pair. Through on-the-fly environment 3D reconstruction, based on visual clues, we propose an integrated method to detect and localise static and moving obstacles, whose position, orientation and speed vector is estimated. Our implementation runs at the moment in a slow real-time (9fps), and should in the future be enclosed in a more complete, probabilistic pipeline.