Intelligence Detection and Identification of Traffic Rule Violations Using a Drone

We propose an intelligence detection and identification method for traffic rule violations by using a drone. First, the proposed method includes a novel approach for detecting solid lanes, bus-only lanes, and shoulder lanes. The second process of the proposed method involves determining each correspondence between detected vehicles and detected license plates. With both the coordinates of the detected lanes and of the vehicle-plate pairs, the next process of the proposed method is an algorithm for identifying vehicles that violate traffic rules. Furthermore, we validate the performance of the proposed method by testing it using video streams taken by a drone in various road environments.