AVP-SLAM: Semantic Visual Mapping and Localization for Autonomous Vehicles in the Parking Lot

FOS: Computer and information sciences Computer Science - Robotics 0209 industrial biotechnology Computer Vision and Pattern Recognition (cs.CV) Computer Science - Computer Vision and Pattern Recognition 02 engineering and technology Robotics (cs.RO)
DOI: 10.48550/arxiv.2007.01813 Publication Date: 2020-01-01
ABSTRACT
Autonomous valet parking is a specific application for autonomous vehicles. In this task, vehicles need to navigate in narrow, crowded and GPS-denied lots. Accurate localization ability of great importance. Traditional visual-based methods suffer from tracking lost due texture-less regions, repeated structures, appearance changes. paper, we exploit robust semantic features build the map localize Semantic contain guide signs, lines, speed bumps, etc, which typically appear Compared with traditional features, these are long-term stable perspective illumination change. We adopt four surround-view cameras increase perception range. Assisting by an IMU (Inertial Measurement Unit) wheel encoders, proposed system generates global visual map. This further used at centimeter level. analyze accuracy recall our compare it against other real experiments. Furthermore, demonstrate practicability application.
SUPPLEMENTAL MATERIAL
Coming soon ....
REFERENCES ()
CITATIONS ()
EXTERNAL LINKS
PlumX Metrics
RECOMMENDATIONS
FAIR ASSESSMENT
Coming soon ....
JUPYTER LAB
Coming soon ....