Home // ICAS 2022, The Eighteenth International Conference on Autonomic and Autonomous Systems // View article

Traffic Signal Recognition and Application Algorithm for the Autonomous Vehicle in V2X Unable Areas

Authors:
Yejin Gu
Daejun Kang

Keywords: safety and traffic efficiency applications; object detection; C-V2X; image processing; CNN; deep learning

Abstract:
Recognizing traffic signals is essential to operate autonomous driving on urban roads. The V2I is the priority in recognizing traffic signals in autonomous vehicles because it can send the exact Signal Phase and Timing (SPaT) of the traffic signal. However, there are many V2I unable areas currently, and even if it is available, it might have trouble due to dommunication delay or fail-operation. Accordingly, we present the traffic signals recognition framework using the convergence of Vehicle-to-Infrastructure(V2I) and camera detection to complement V2I. This study uses the signal recognized by the camera sensor in the area when the V2I signal is unable. By changing the existing one-way decision-making method, we implemented a customized communication system that dynamically changes in real time and fits the infrastructure situation. We identified that the proposed method works well by deploying an autonomous driving pilot system on the designated segment in Sejong-si, South Korea, where the V2I is partially available.

Pages: 14 to 18

Copyright: Copyright (c) IARIA, 2022

Publication date: May 22, 2022

Published in: conference

ISSN: 2308-3913

ISBN: 978-1-61208-966-9

Location: Venice, Italy

Dates: from May 22, 2022 to May 26, 2022