Abstract:
The project aimed to develop an automatic system for speed monitoring and number plate
detection of vehicles using a drone. The system utilized two key technologies: deep neural
network (DNN) for number plate detection and image processing for speed detection. The
DNN was trained on a large dataset of vehicle images to detect number plates with high
accuracy, and the system used a pre-trained DNN model to identify the region of interest
containing the number plate. The image processing component of the system was designed
to detect the speed of the vehicles in real-time using a video camera mounted on a drone.
The captured images were processed using optical flow techniques to detect the movement
of vehicles, and a threshold based approach was used to identify vehicles moving above a
certain speed limit. The system was tested in real-world scenarios and demonstrated high
accuracy in detecting and capturing number plates of vehicles moving above a certain
speed limit and the ability to detect the speed of moving vehicles in real-time. This project
showcases the potential of combining deep neural network and image processing
techniques to develop an effective and feasible system for automatic speed monitoring and
number plate detection, with possible applications in traffic monitoring and law
enforcement. Further optimizations and improvements can be made to enhance the
system's performance and functionality
