On-Line Arabic Handwriting Text Line Detection Using Dynamic Programming
Document Type
Conference Proceeding
Publication Date
2011
Abstract
Text line detection in unconstrained handwritten documents remains an open document analysis problem. The typical method based on horizontal projection analysis and regrouping connected components is not usually successful because of baseline undulations and shifts, baseline-skew variability and inter-line distance variability.This work deals with on-line Arabic handwritten document segmentation, such as pages of handwritten notes. We propose an automatic text lines detection method based on dynamic programming. We try to find the paths with the minimum cost between collections of text lines segments. Most steps of the proposed algorithm are based on off-line information. Hence it can also be applied to off-line documents after a few minor changes.The proposed methodology is tested on OHASD the first Arabic online sentence dataset. More than a hundred handwritten Arabic texts written by different writers are examined. Our experiments show better results than common on-line segmentation procedures .
Recommended Citation
Elanwar, Randa Dr.; Rashwan, Mohsen A. Prof.; and Mashali, Samia Prof., "On-Line Arabic Handwriting Text Line Detection Using Dynamic Programming" (2011). Computer Science. 105.
https://buescholar.bue.edu.eg/comp_sci/105