Vision Based Localization for Infrastructure Enabled Autonomy

摘要: Infrastructure Enabled Autonomy (IEA) is a new paradigm in autonomous vehicles research that aims at distributed intelligence architecture by transferring the core functionalities of sensing and localization to infrastructure. This also promising designing scalable systems enable car platooning on highways. paper gives detailed description about experimental realization IEA techniques devised localize vehicle such setup. A reliable camera calibration technique for an setup discussed, followed transform 2D image coordinates 3D world coordinates. In this research, information received from on-board sensors like GPS/IMU, (2) localized position data derived deep learning, coordinate transformations real-time feeds (3) lane detection infrastructure cameras. fused together utilizing Extended Kalman Filter (EKF) obtain estimates 50 Hz. then used control with objective following prescribed path. Extensive simulation results are presented corroborate performance proposed approach.