Traffic Light Detection System in Self-Driving Cars

Abstract

In recent years, Self-driving cars have been a topic of interest. These self-driving cars can optimize their driving behavior to maximize fuel efficiency and reduce traffic congestion. The paper discusses the challenges of traffic light detection in self-driving vehicles and suggests ways to address them. Traffic light detection is a critical component of self-driving car technology, as it enables autonomous vehicles to accurately perceive and respond to traffic signals. An overview of traffic light detection is provided in self-driving cars, outlining its benefits and drawbacks. It reviews the existing literature on traffic light detection and highlights various approaches, including computer vision methods, machine learning techniques, and deep learning models that have been proposed to address this issue. The algorithms like SVM, Heuristic and CNN algorithms are proposed in the existing solutions. These algorithms work accurate during day time but not in night time. And also, by using many different sensors for detection the main drawback is that they can fail over time which may lead to costly replacements. This study examined various traffic light detecting algorithms and their procedures. Also note the difficulties and upcoming work on traffic signal detection for precise findings. SVM is frequently employed, and CNN and Faster RCNN are well-known algorithms for object detection and traffic signal color categorization.