Open AccessArabic Handwritten Recognition by using His Features
Author(s) -
Ali Mirza Mahmood,
Sabah Kadem
Publication year - 2021
Publication title -
magallaẗ kulliyyaẗ al-rāfidayn al-ǧāmi'aẗ al-'ulūm/maǧallaẗ kulliyyaẗ al-rāfidayn al-ǧāmiʻaẗ li-l-ʻulūm
Language(s) - English
Resource type - Journals
eISSN - 2790-2293
pISSN - 1681-6870
DOI - 10.55562/jrucs.v26i1.428
Subject(s) - cursive , computer science , artificial intelligence , pattern recognition (psychology) , segmentation , tracing , intelligent character recognition , intelligent word recognition , character (mathematics) , feature (linguistics) , speech recognition , chain code , document processing , smoothing , computer vision , character recognition , image (mathematics) , mathematics , linguistics , philosophy , geometry , operating system
The present study is a contribution in the filed of Arabic character recognition. The first step is to enter an image from a scanner, and then some special processes are done on the image like (smoothing and thinning and resizing) before continuous to segmentation phase. Then modified moment invariant computed using the shape boundary of the character only is utilized as features.Also this study is concerned with the recognition of handwritten Arabic character drawn on a graphic tablet. At first feature that are suitable for recognition are proposed. Feature that are found to be independent of the writer style are represented as a list (vector) of integer values.The designed system in general view is an Arabic text recognition system basically includes a segmentation stage in order to recognize handwritten Arabic words. The segmentation process is completely based on tracing the outer of given word and calculating the output distance between the extreme points of intersection of the contour with vertical line. At the output of the segmentation stage the cursive word is presented as a sequence of isolated character contour.