A robust and effective character recognition approach is the first stage of providing automatic access and sophisticated retrieval functionalities for digital Ottoman archieves. In this study, a novel segmentation and recognition method for connected letters of Ottoman script is presented. In this method, all possible segments of a connected component are generated vertically and thereafter the candidate letters to which extracted segments are most similar are determined. Next, a function is defined for scoring each distinct syntactically correct sequence of these candidate letters, whose segments spans the connected component. In order to find the candidate letter sequence that maximizes the score function, a directed acyclic graph is constructed. The set of letters residing on the connected component are finally determined by finding the longest path in this graph. Experiments using a collection of printed Ottoman documents reveal that the proposed method provides very high precision and recall figures.

DATE: 5 May, 2008, Monday@ 16:15
PLACE: EA 409