Text Line detection and Segmentation in Handwritten Gurumukhi Scripts

Gurumukhi script is a two dimensional composition of symbols with connected and disconnected diacritics. Handwritten Gurumukhi script has some complexities like connected, overlapped text lines. It is one of the major reasons for errors during the recognition process. Text line segmentation is a challenging job in unconstrained writer independent handwritten document image processing. There is a great need for research in the area of Punjabi handwriting recognition to resolve challenging problems involved in it. In this paper we have proposed an algorithm for text line segmentation in handwritten Punjabi document that deals with the problems like overlapped and connected components in text line and extract text lines from handwritten document image. The text line detection algorithm is based on locating the most favourable segments of text line and associating it with its respective text line inserting a gap between neighbouring text lines. Keywords— Text line segmentation, overlapped text lines, connected text lines, average height, Gurumukhi script