Real-Time Speed Estimation, License Plate Recognition of Detected Vehicles for the Intelligent Traffic Management System
摘要
This paper presents a system that works to spot the moving vehicles on figure out speeds, and read license plates in real time scenario simultaneously. The aim of this work is to help in roads safety and managing traffic. This system uses the Faster R-CNN framework with a ResNet-50 base to find vehicles in different situations. It also uses EasyOCR to read letters and numbers on license plates, which helps identify vehicles. The system estimates over speeding vehicles by using dlib’s correlation tracker and changing pixels to meters. This helps enforce traffic rules. Our system is built to work with traffic management setups. The customization is also possible in this system. This paper also shows the deep learning utilization in traffic systems.