Reliable Autonomous Driving Environment Model With Unified State-Extended Boundary

From the early stage of robotic applications to current autonomous driving technologies, environment modeling has been acting as middleware for connecting perception and decision layers. In applications, space-oriented models (e.g., grid map, drivable area) are widely applied faithfully reflect space occupation. With development driving, highly dynamic complex road brings rising need understand type motion status objects, thus element list became mainstream model. However, along comes reliablity problem caused by missed detection irregular which is still inevitable despite accuracy improvement. view this, a new proposed unified state-extended boundary (USEB), aiming improve element-oriented For requirements, different types elements consistently converted into constraints. Semantics dynamics expressed area boundary, making it possible merge occupation reliability against objects. Evaluation USEB carried out on nuScenes dataset. Comparative results show that could cover required information decision, whereas achieving higher than commonly