TY  - JOUR
T1  - Effective Method for Segmentation of Arabic Manuscript Documents
AU - Aboubekrine, Aicha Mint AU - KADIRI, Kamal Eddine EL AU - Tabii, Youness AU - Diakite, Mohamed Lamine AU - Mohammed, Lamarti Sefian AU - Taleb, Nagi ould 
JO  - International Journal of Soft Computing
VL  - 14
IS  - 1
SP  - 28
EP  - 32
PY  - 2019
DA  - 2001/08/19
SN  - 1816-9503
DO  - ijscomp.2019.28.32
UR  - https://makhillpublications.co/view-article.php?doi=ijscomp.2019.28.32
KW  - Arabic handwritten documents
KW  -vertical projection profile
KW  -horizontal projection profile
KW  -binary
image
KW  -Arabic manuscript
KW  -segmentation
AB  - In this study, we are interested in Arabic handwritten documents. We propose a method of
segmentation of Arabic manuscript documents first in lines, then in words and finally, proceed to their
annotation. This research poses a number of problems because of the difficulties related to the processing of
Arabic manuscripts. After binarizing the images of the Arabic handwritten documents, we compute the vertical
projection profile of the text image to determine the lines of text, then the horizontal projection profile for the
segmentation of lines into words and finally, the detection of under words and real words. The technique we
offer allows us to access the content of the handwritten documents in a very efficient and fast way.
ER  -