AUTOMATIC MULTI-STYLE EGYPTIAN LICENSE PLATE DETECTION AND CLASSIFICATION IN HIGH-RESOLUTION VIDEOS USING DEEP LEARNING METHODS

Abstract

This thesis presents state of art of deep learning methods to detect and classify Multi-style Egyptian car license using state of the art Yolo object detection followed by Siamese network or multi-task learning methods for classifying the most four common different classes of Egyptian license plates with its two main categories standard and non-standard Egyptian license plates.