A texture based approach for Arabic writer identification and verification

In this paper, we present a novel approach for Arabic Text-Independent Writer Identification and Verification. Given that the handwriting of different people is often visually distinctive, we propose a global approach based on texture analysis, where each writer's handwriting is regarded as a different texture. This allows us to apply a texture classification method mainly based on a set of new proposed features extracted from Grey Level Run Length (GLRL) Matrices. The efficiency of the proposed approach is demonstrated experimentally by the classification of 650 handwriting documents collected from 130 different Arabic writers. Comparisons with Grey Level Co-occurrence Matrices (GLCM) technique demonstrate that the GLRL matrices contain more discriminatory information and that a good method of extracting such information is of great importance for successful classification.