Abstract:Locating and classifying different types of vehicles is a vital element in numerous applications of automation and intelligent systems ranging from traffic surveillance to vehicle identification, with deep learning models now dominating the field of vehicle detection. However, vehicle detection in Bangladesh remains a relatively unexplored research lacuna. One of the main goals of vehicle detection is its real-time application, with “You Only Look Once” (YOLO) models proving to be the most effective. This paper compared real-time vehicle highway detection systems using YOLOv4, Faster R-CNN and SSD algorithms to determine the best performance. A vehicle detection and tracking system was also developed that improved highway safety. Vehicle trials compared the real-time performances of the YOLO, Faster R-CNN and SSD algorithms in detecting and tracking highway vehicles by measuring precision, recall, F1-score and operating speed. Models for each algorithm were constructed and each model was trained and tested, with performance measured using a confusion matrix. This statistical tool assessed the efficiency of the system using a prepared test dataset and evaluated the results using appropriate indicators such as real-time road lines, traffic signs and vehicle detection false positive rates. Results showed that the YOLOv4 algorithm outperformed Faster R-CNN and SSD in real-time vehicle detection and tracking on highways. YOLOv4 also processed the results more quickly and proved superior in detecting and tracking objects in real time. The Faster R-CNN algorithm gave high object detection, tracking accuracy and recall while reducing the number of locations needing detection, with the SSD algorithm providing high precision, recall and good image detection results.

Detection of Passing Vehicle on the Highway

Vehicles Detection and Tracking Based on the Sub-Feature

A Robust Video Based Traffic Light Detection Algorithm for Intelligent Vehicles

Vision‐based Inter‐vehicle Distance Estimation for Driver Alarm System

Detection of engineering vehicles in high-resolution monitoring images

Embedded and Real-Time Vehicle Detection System for Challenging On-Road Scenes

Real-time Vehicle Detection and Distance Measurement in Multi-Environment

Real-time vehicle detection and tracking based one traffic scene analysis

Moving Object Detection Using an In-Vehicle Fish-Eye Camera

Real-time Structured Lane Detection by Monocular Vision

Detection of overtaking vehicles in a highway

A fast detection algorithm for moving vehicle in traffic scenes

Robust Lane-Detection Method for Low-Speed Environments

Moving Vehicle Detection for Automatic Traffic Monitoring

Efficient Lane Detection System Based on Monocular Camera.

Detection and recognition method of monocular vision traffic safety information for intelligent vehicles

A Real-Time Drivable Road Detection Algorithm In Urban Traffic Environment

Real-time vehicle detection system on the highway

An Algorithm for Front-vehicle Detection Based on Haar-like Feature

Detection of highway lane lines and drivable regions based on dynamic image enhancement algorithm under unfavorable vision

Real-time multiple vehicles tracking with occlusion handling