A higher percentage of people are riding motorcycles in emerging nations like India. Automated helmet law enforcement systems can dramatically improve traffic safety. The goal of this work is to create a reliable Number Plate Recognition (NPR) system that is specifically designed to identify and sanction motorcycle helmet infractions using an Improved Faster R-CNN method. Modern object identification models are used in the system's initial step of multi-stage license plate detection. Character segmentation and recognition techniques are used to obtain registration numbers after license plate identification. In order to categorize helmets using CNN, a number of models of the technology were utilized, however most of them needed an image preprocessing step to extract the Region of Interest (RoI) region from the picture. The system's integration into existing traffic surveillance infrastructure can aid law enforcement agencies in efficiently identifying and addressing helmet-related violation, thereby contributing to improved road safety and reduced head injuries in motorcycle accidents.