Litcius/Paper detail

Traffic Management System Using YOLO Algorithm

Pankaj Kunekar, Yogita Sachin Narule, Richa Mahajan, Shantanu Mandlapure, Eshan Mehendale, Yashashri Meshram

202417 citationsDOIOpen Access PDF

Abstract

The issue of traffic congestion is becoming worse day by day. The typical traffic lights are unable to effectively regulate the growing number of vehicular traffic; therefore, we mixed computer vision and machine learning to mimic complicated incoming traffic at signalized intersections. This was accomplished using the cutting-edge, real-time object detection system You Only Look Once (YOLO), which is built on deep convolutional neural networks. In order to maximize the number of vehicles that can cross safely with the least amount of waiting time, this paper presents an efficient method to use this algorithm, where traffic signal phases are based on the data obtained, primarily queue density and waiting time per vehicle. Embedded controllers that adopt the transfer learning methodology can implement YOLO.

Topics & Concepts

Computer scienceQueueReal-time computingConvolutional neural networkEnhanced Data Rates for GSM EvolutionTraffic congestionTransfer of learningDeep learningFloating car dataTransfer (computing)Artificial intelligenceAlgorithmComputer networkTransport engineeringEngineeringParallel computingAutonomous Vehicle Technology and SafetyAdvanced Neural Network ApplicationsTraffic Prediction and Management Techniques