학술논문

Lane Detection With a High-Resolution Automotive Radar by Introducing a New Type of Road Marking
Document Type
Periodical
Source
IEEE Transactions on Intelligent Transportation Systems IEEE Trans. Intell. Transport. Syst. Intelligent Transportation Systems, IEEE Transactions on. 20(7):2430-2447 Jul, 2019
Subject
Transportation
Aerospace
Communication, Networking and Broadcast Technologies
Computing and Processing
Robotics and Control Systems
Signal Processing and Analysis
Sensors
Radar cross-sections
Roads
Radar detection
Laser radar
Automotive engineering
autonomous driving
automotive radar
clustering
lane detection
road marking
radar sensor
radar cross section
radar reflector
Language
ISSN
1524-9050
1558-0016
Abstract
Autonomous driving will significantly shape the near future of transportation that requires the distinct knowledge of the driving environment, especially the lane boundaries ahead of the vehicle. For this sensing task, optical and automotive radar sensors are mostly applied, while the radar sensor is less sensitive to non-ideal illumination conditions. Road boundaries can be detected by radar sensors through objects like guardrails, delineators, road curbs, and so on. However, in a multi-lane roadway or with missing roadside infrastructure, the adjacent lanes and the road boundary also have to be recognized by the radar sensors. This makes it necessary to integrate appropriate radar reflectors into the lane or road boundaries. Such reflectors, when integrated in the road, should not harm the vehicle tires, so their height shall be low, but they still have to be able to reflect the radar signals properly. Conventional road markings can be accordingly adjusted with such reflectors to support vehicle guidance with radar sensors. In this paper, the scattering property of various types of reflectors is evaluated by simulations over a wide angular and range geometry. The simulation results are analyzed with regard to the view angle influence on the radar cross section. Then, the specimen measurement results will be presented and finally the clustering process will be introduced.