With increase in the population and improving lifestyle, the demand for transport has increased
tremendously. Increase in number of vehicles, increases the risk of accidents. Automatic helmet
detection plays an important role in preventing the people from breaking the rule of “Helmet
Wearing”. Many companies / institutions have, therefore, imposed a rule which says “No entry
without helmet”. In this report, we propose an approach for automatic gate opening for two-wheeler
riders wearing helmets. The proposed approach detects two-wheeler riders from surveillance video and
then it determines whether the two-wheeler rider is wearing a helmet or not. Transfer learning has
been implemented using Yolo (You only look once) for the purpose of detecting the two-wheeler
riders with and without helmets in the real-time with the help of surveillance cameras in order to
determine whether the gate needs to be opened or not.