License plate recognition

Abstract

In today’s ever-growing technological society, Automatic License plate Recognition, ALPR, has many implications for solving traffic-related applications and transporta- tion planning. Identifying cars in pursuit or stolen cars, controlling automatic park- ing access, registering missing vehicles from last found footage, and in many more hazardous or unpredictable situations, ALPR helps to identify and extract license plate information from surveillance footage. Thus in improving and making ALPR efficient, many techniques have been introduced with algorithms playing an essential part for vehicle surveillance systems, although many challenges are seen in correctly computing and recognizing license plates under different environmental conditions. In this research, we work with different algorithms for understanding Bangladeshi license plates, analyze the algorithms’ efficiency in various environmental conditions or unlikely situations, and compare them with our model, which currently is giving 97% accuracy, to find the most suitable for recognizing them.