TrafficNN

Abstract

Traffic congestion can affect different socio-economic aspects of a country. In order to perform better than traditional human effort in traffic control, an automated system must be developed. For reducing the traffic congestion, traffic conditions classification can be considered as the first way to monitor the traffic control system. So, detection of traffic conditions is crucial for building a smart traffic control system to prevent traffic jam escalation. Deep neural network is the area of machine learning that is being widely used for interpreting and analyzing visual data. This branch has an extensive use over image classification. So, modern deep learning approaches can help in detection of road traffic conditions especially Convolutional Neural Network (CNN). We suggest a novel model in this paper that can classify road traffic conditions using CNN. Our proposed model ‘TrafficNN’ classifies five different road traffic conditions with an accuracy of 82%. To train and test our model, we use our own traffic conditions images dataset. To verify the efficiency of our model, we compare it with several pre-trained models like- VGG16, ResNet50, InceptionV3 and DenseNet121. The comparison result proves the significance of our model to extend its successful application for developing an automated traffic controlling system in the near future.