Arabic words recognition technique for pattern matching using SIFT, SURF and ORB

• Main Authors: Mohd Zailani, Syarah Munirah, Morshidi, Malik Arman, Mohd Esa, Luqman Naim
• Format: Article
• Language: English
• Published: Universiti Teknikal Malaysia Melaka 2017
• Subjects: QA75 Electronic computers. Computer science; QA76 Computer software
• Online Access: http://irep.iium.edu.my/60355/1/1570378451.pdf; http://irep.iium.edu.my/60355/
• Institution: Universiti Islam Antarabangsa Malaysia

Image matching technique requires a robust and fast technique to be applicable in various application. This paper investigates which recognition technique suits better in matching an image of printed Arabic text. The recognition algorithm involves the conventional Scale-Invariant Feature Transform (SIFT), Speeded Up Robust Features (SURF) and Oriented FAST and Rotated BRIEF (ORB). A parameters estimator of models algorithm is used to weed out the outlier point of matching images. The test demonstrates on the Arabic word images with the different angles, scales, and viewpoints. We evaluate the performance through analyzing the matching accuracy rate and computational time