Document Type : Research Article

Authors

• D. Palanikkumar ¹
• Abhisek Sethy ²
• Kasiprasad Mannepalli ³
• P. John Augustine ⁴
• Semira Abdela Seyid ⁵
• Adem Aman Hassen ⁶

¹ Professor, Department of IT, Dr NGP Institute of Technology, Coimbatore, India.

² Assistant Professor, Department of Information Technology, GMR Institute of Technology, Rajam, Andhra Pradesh, India.

³ Associate Professor, Department of ECE, K L Deemed to be University, Vaddeswaram, Guntur, Vijayawada, Andhra Pradesh, India

⁴ Associate Professor, Department of Computer Science and Engineering, Sri Eshwar College of Engineering, Coimbatore, Tamil Nadu, India.

⁵ Lecturer, Department of Mechanical Engineering, Dambi Dollo University, Dambi Dollo, Oromia Region, Ethiopia.

⁶ Department of Mathematics, College of Natural and Computational Science, Dambi Dollo University, Dambi Dollo, Oromia Region, Ethiopia.

Abstract

Much traditional software is available for recognising the handwritten characters or digits and that gets converted to the digital format. But the accuracy level and performance of the systems are not that much acceptable due to lack of temporal information. In this proposed method the training set is created with the strokes and segments of the characters and the digits and that get matched with the test data. The segmentation is based on position of character, order of writing the character, writing time and writing pressure. A mathematical model is created with a support vector function to perform the identification. The stroke and segment extraction makes the accuracy more in this system. The aim is to build a model for managing handwritten text recognition accuracy and translate them into speech. This problem is approached using Recurrent Neural Network.

Keywords

• Text Recognition
• Accuracy
• neural network