Automatic Number Plate Recognition

Automatic Number Plate Recognition (ANPR) systems have gained a lot attention due to its vast applications, including traffic management, law enforcement, toll collection and parking systems. ANPR can also be used to track stolen vehicles that have been used in criminal activities.
ANPR systems face challenges in dusty environments, especially in developing countries where there are many untarred roads. ANPR systems also face tedious task due to the variety of plate formats (i.e., Plate size, background, character size, plate texture) especially where there is a busy environment and complex scenarios. To address some of these challenges we present license plate dataset captured at various locations in Ghana.
The dataset is presented into one main folder which contains twelve (12) sub folders. The total number of images in the dataset is 11,326 images. It is grouped according to the regions in Ghana (6,202-Ashanti Region, 440-Brong Ahafo Region, 103 -Central Region, 665 - DV, 70 - Eastern Region, 3,084 -Greater Accra Region, 559 -Motorcycle, 24 - Northern Region, 10 -Upper East Region, 19 -Upper West Region, 07 - Volta Region, and 143 -Western Region).
This dataset will contribute to the development of a more robust and efficient Number Plate Recognition system.