Helmet detection and license plate recognition

Motorcycles are widely used in developing countries as the essential mode of transport. In last recent years, there has been a rapid increase in road accidents owing to the fact that majority of the motor bicyclist fail to wear helmet that makes it an ever-present danger. In this decade, most of the accidents are caused because of the head injury. Due to this, wearing helmet is formed necessary by means of traffic rules. But most of the motorbike riders never obey the rule. Many cities make use of a surveillance network to monitor the bicyclists violating the helmet laws. But such a system will need human intervention. The surveys say that human interventions prove ineffective, due to the increase in the time of monitoring and also due to the errors made by human during monitoring. The main idea of this application is to detect the license plate number of those who violate the law of wearing helmet using Deep Learning Algorithms. To detect the bicyclists who are against the helmet laws, a system which uses convolutional neural network and image processing is implemented. This system deals with the detection of motorbike, Helmet or Non-Helmet classification and License plate recognition of Motorbike. YOLOv3 feature is used to detect the motorbikes. When the bike is detected using convolutional neural network(CNN), it determines whether the bicyclist wears a helmet or not. If the rider is identified with no helmet, then the license plate number of his motorcycle is captured and extracted using Tesseract OCR.