On-The-Fly Traffic Sign Image Labeling

With the rapid development of society and economy, automobiles have now become the most convenient mode of transportation. It is mainly the use of vehicle cameras to capture real time images of roads and to track and identify road signs. Therefore, providing the driving device with precise information. The Dataset that is being used are a collection of traffic signs. Using the darknet module we set up YOLO(You Only Look Once). Yolo is a predefined CNN which is used to label objects automatically. YOLO works on a dataset called the COCO dataset which consists of data from 80 different classes. However YOLO is trained with data from the US, as a result of which it won't be able to detect objects in India. Hence, a new CNN has to be trained which is capable of detecting signs back here in India.