Vehicle license plate recognition systems / Imran Ismail

• Main Author: Ismail, Imran
• Format: Thesis
• Language: English
• Published: 2004
• Subjects: Automobile license plates; Optical data processing
• Online Access: https://ir.uitm.edu.my/id/eprint/68635/1/68635.pdf; https://ir.uitm.edu.my/id/eprint/68635/

The purpose of the vehicle license plate recognition project is to recognize or identify vehicles. Vehicle license plates are meant to be recognized by people, especially government officials such as policemen and transport personnel. In current situation, we can identify all the information's related to the drivers from number plates. Vehicle's license plates that are authorized and registered are allowed to enter the campus premises. In this project, a database of vehicle license plate numbers that are authorized to enter the university grounds has been compiled. The project started with identification of a vehicle consists of three major parts namely sensing, image capturing and recognition of the license plate. A car is sensed and a signal is passed to the image-capturing receiver and triggers the video camera to take a snap using a high-speed shutter to reduce motion blur. Before image recognition takes place, the image needs to undergo pre-processing. This includes image cropping, image quantization and filtering and character segmentation. Only with all these image processing can the recognition of the numerals take place efficiently minimizing errors. In this project I have decided to choose neural network to recognize each character among other types of character recognition systems. This neural network has been trained off-line with a set of characters.