Real Time Image Classification Based Smart Railway Platform Using Convolutional Neural Network

By supporting smart cities and smart railways, India is heading in a smart direction now-a-days, we have taken a small initiative in this aspect by introducing an innovation in the Indian railway system in the domestic railway platform. The goal for this innovation is to resolve few issues faced by physically disabled individuals and senior citizens while crossing the bridge between the two railway platforms by foot. A updated smart platform is suggested on paper to solve this problem. The Smart Platform on the side of the Crossing Point Platform is interested in the railway tracks. The Smart Platform will open when there is no train arriving at the station. The smart interface can be used for visually disabled people to cross between platforms. On both sides of the road, proximity sensors and the camera unit are configured. If the train hits one sensor, the signals will be transmitted to the camera unit, the image will be captured further by the camera, and the CNN algorithm will decide whether the image captured is some other image of the train. The virtual platform works on the basis of this performance. If the captured image is of a train, the smart platform will then close and allow the trains to go through the tracks, and the smart platform will open automatically between the platforms when the train leaves the second sensor. The train entering the railway stations will be announced and demonstrated by LED signals in a voice system. The proposed system provides the best response without using foot over bridge for the travel of physically disabled people from one platform to another platform.