Analysis and Recognition of Bilingual Handwritten Scripts
Panyam Narahari Sastry1, G. Akhil2, Vaishnavi Suthram3
1Dr. Panyam Narahari Sastry, Professor, Department of ECE, Chaitanya Bharathi Institute of Technology (A) ,Hyderabad, India.
2G. Akhil, UG, Department of ECE, Chaitanya Bharathi Institute of Technology (A) ,Hyderabad, India.
3Vaishnavi Suthram, UG, Department of ECE, Chaitanya Bharathi Institute of Technology (A), Hyderabad, India.
Manuscript received on March 16, 2020. | Revised Manuscript received on March 24, 2020. | Manuscript published on March 30, 2020. | PP: 2571-2578 | Volume-8 Issue-6, March 2020. | Retrieval Number: F8538038620/2020©BEIESP | DOI: 10.35940/ijrte.F8538.038620
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In this work, offline handwritten character recognition (HWCR) is involved, which is an open area of research for Indian languages. The recognition accuracy for HWCR is around 60% as per the literature survey. The main obstacle for the research in this area is the non-availability of a standard database. Character Recognition (CR) is an application of pattern recognition. Pattern recognition has many applications like security services, defense organizations, banking, post offices, archeological field, weather forecasting, library automation, reading aids for the visually challenged, etc. There are very less number of users for Indian languages when compared to English and hence the research for HWCR is at early stage. In this work, transform based recognition techniques are used on two languages namely Hindi and English. The best recognition accuracy obtained for the bilingual handwritten scripts is 73.33% which is in line with the existing research publications.
Keywords: 2D-FFT, 2D Correlation, Bilingual Characters, Nearest Neighborhood Classifier, Optical Character Recognition (OCR).
Scope of the Article: Measurement & Performance Analysis.