3D camera-based perception platform
3D obstacle detection
Autonomous transport systems have to overcome two major challenges: on the one hand, collision avoidance with objects and persons, on the other hand, autonomous avoidance of obstacles. The frequently used safety scanners are only of limited help here, as they only detect the travel path in a plane just above the ground. This is where the camera platform shows its advantage: it processes the signals from up to six 3D PMD cameras installed all around the vehicle and evaluates the environment three-dimensionally, i.e. both the ground area below the field of view of the safety scanners (e.g. holes in the ground) and the view diagonally upwards. In this way, hanging loads such as crane hooks, for example, are also detected. Powerful algorithms ensure that false detections are virtually eliminated despite the high detection rate.
Easy integration
The user can define zones in the form of segmented polygons in which the system evaluates the occupancy and provides the vehicle’s steering system with clear data for safe and collision-free driving.
Picture 1: The obstacle in front of the vehicle is projected in a map on the ground. One of the three zones or the area in the so-called occupancy grid is then output as occupied. Picture 2: The robotics platform captures the situation in a 2D image and in 3D distance data.
