A database of Arabic handwritten characters

Arabic character recognition is an important aspect in pattern recognition because there are many applications that depend on it. Database availability is a key factor that encourages researchers to conduct research in this field. This paper proposed a database with a variety of shapes for each character. The database contains 5,600 images written by 50 writers. Each character has 200 images divided into four groups (beginning, medium, end and isolated). This database can be used by the machine learning community for training and testing, in general, the database will help the researchers to compare and evaluate their algorithm performance.