Detecting Helmets Of The Bike Riders Using Deep Learning Algorithms

Abstract

We propose real-time bike helmet detection in our study. A lot of people ride bikes in our nation. Motorbikes are more popular than vehicles because they are cheaper to maintain, take up less parking space, and provide more mobility and adaptability in urban situations. Bike riding is entertaining yet risky. Bicyclist safety is the planned system's main purpose. Many drivers don't wear helmets even though they're mandated by law. In emerging countries, mortality has been growing steadily. A helmet detection technology that identifies drivers without helmets is needed to safeguard the public. We employ a real-time 3202 dataset for this approach. We gather wearing helmet 1911 and no helmet 1291 data and utilize algorithms like VGG16, Resnet50, MobileNet v.02, Inception V3, EfficientNet, and CNN. The EfficientNet achieved 98% accuracy. Each person's comparison statement techniques are in the implementation section. To construct the optimum model for the conditions, this inquiry uses model validation approaches.