Neural Network Based Traffic Monitoring using UAVs

Abstract

In large and growing metropolitan areas, the rise in traffic congestion is becoming an inescapable problem. It is estimated that the traffic congestion in metro cities costs the nation approximately 1.5 lakh crore rupees every year. With the increase in congestion, accident rate increases proportionally. The reckless driving and increased speed are the root cause of road accidents. We propose a speed detection algorithm to detect and monitor the speed of vehicles crossing a certain threshold speed limit. On national highways, the long queues at toll booths lead to loss of time and money. We propose image processing and convolutional neural network based algorithm to address the problem of traffic congestion, ease the flow of traffic, anomalies detection and ultimately reduce pollution and fuel consumption.