Vol.1 No.4
Year : 2014
Issue : Oct-Dec
Title : Unconstrained Handwritten Text Line Segmentation: An Improved Technique
Author Name : kiran, Mohammadi , Manjunath Aradhya
Synopsis :
Handwritten text line segmentation is still a challenging area in document image analysis. This is due to the variation of handwritten style from individual to individual causing many problems like multi-skewed, overlapping and touching text lines. In this paper, authors have proposed an improved technique for handwritten text line segmentation to the method proposed in [1]. The proposed method consists of four major phases, which includes piece-wise painting algorithm, skew estimation, line drawing and edge detection process. The proposed method is implemented on dataset of Kannada, English and comprising of Persian, Oriya and Bangla documents. The success rate of the technique, as observed experimentally was satisfactory and it may contribute significantly for the development of applications related to handwritten OCR (Optical Character Recognition).
For more details:
