Title:	SEGMENTATION AND RECOGNITION OF CONNECTED HANDWRITTEN CHARACTERS
Author(s):	Shehan Panditharatne and Lochandaka Ranathunga
ISBN:	978-989-8704-21-4
Editors:	Yingcai Xiao, Ajith P. Abraham and Jörg Roth
Year:	2020
Edition:	Single
Keywords:	Implicit, Explicit Segmentation, Connected, Touching Characters
Type:	Full
First Page:	63
Last Page:	70
Language:	English
Cover:
Full Contents:	click to dowload
Paper Abstract:	Optical Character Recognition is still an open problem in computer science. This paper explores segmentation of connected (touching/overlapping) handwritten characters through implicit segmentation and explicit segmentation. Explicit segmentation is achieved through rules based on background thinning and analyzing the contour hierarchy. Implicit segmentation is achieved through generating a dataset of connected characters and training a CNN. This research was inspired by the need to recognize math documents containing algebraic equations. Recognition of 72 characters (alphanumeric and math symbols) are achieved through the proposed methods. The scope of the implemented OCR is limited to maximum of two connected characters.
	Go Back