Host: The Japan Society of Mechanical Engineers
Name : [in Japanese]
Date : June 05, 2019 - June 08, 2019
This paper proposes an experimental result about real-time recognition of the position and the orientation of traffic lights based on this Photo-model-based method in a simulated environment. Because experimenting at real traffic intersections can cause a lot of trouble for pedestrians and vehicles, we have built a road simulation environment. The Photo-model-based method can obtain the current position and posture of the object by using only two calibrated cameras through the image of the object that has been acquired. I will introduce the fitness function of this method and the method of exploration using genetic algorithms in this article, as well as the latest data in this environment. This time, the results of the experiment were obtained by experimenting with traffic lights in red states. The experimental results in real time will be reported.
