The contemporary world has seen a radical increase in large congestion of 

automobiles at toll-booths. In the preceding years manual collection was widely practiced 

across India. For this system to work uneventfully, we require a huge workforce as the 

situation needs to be monitored perpetually. The toll collector decides the amount to be 

paid upon the type of the vehicle, its features, journey type and pre-determined monetary 

values set by the government. So as to overthrow the hitches met in the manual system,

Electronic Toll Collection (ETC) came into existence. The primary objectives of such 

systems is that to allow clearance for vehicles with minimal setback through the toll plaza.

The objective of this project is by using complicated features of a vehicle and 

decoded with the help of computer vision where each constraint can be used to differentiate 

and analyze them. The system mainly consist of two modules for input: vehicle detection, 

vehicle type recognition and secondly number plate recognition. Both of the modules take 

place parallel and separately .Vehicle recognition system has indeed become a vital part of 

any intelligent or smart transportation system. In our system we use YOLO, an algorithm 

which is built upon Convolutional Neural Network (CNN) and deep learning and also the 

second module is number plate recognition, Opencv and a machine learning model KNN 

(K near neighbor). Finally a user interface is built for viewing vehicle entry data by toll 

gate officials

Share with your friends

Team Members


Post Your Comment

Please Login to Comment