Road Traffic Condition Monitoring using Deep Learning

Abstract

Massive amounts of information about vehicular traffic are added to the traffic monitoring system every second. Keeping an eye on these metrics demands a lot of time and labor. It is possible to use a deep learning method (Convolutional Neural Network) for traffic management and control. The training dataset is built using pre- processed traffic monitoring data.
Constructing the Traffic net entails retraining the network on data from traffic-related applications and shifting the network to the new domain. There are several large-scale uses for this Traffic net, one of which is regional detection. More importantly, it has widespread applicability. Faster finding and higher precision in the case study are impressive proof of efficiency. The preliminary evaluation may lead to its implementation in a traffic monitoring system and may eventually improve the intelligent transportation system.