Smart Traffic Surveillance System

The use of helmets for riding motorbikes is critical and vital as it helps prevent life-threatening injuries if an accident occurs. Helmet detection is a significant yet challenging vision task in traffic monitoring and accidents avoidance for bike riders. With the increase in bike accidents in several nations due to the lack of power to impose traffic laws and mostly the negligence of the bike rider. The proposed method aims to detect no helmet bikers and extract the license number plate so that they can be issued a fine. This article highlights an automatic helmet detection method, which incorporates, Feature descriptor and DNN-classifier. The system is trained on predefined datasets and uses the DNN (Deep Neural Networks) classifier to classify the detected vehicles, as described in the proposed method works as follows, Preprocessing, Feature extraction and Classification. This article proposed a method that will classify whether or not the person riding the bike is wearing a helmet. In terms of robustness and efficiency, this proposed method achieves high accuracy detection rates.