Dhivehi OCR Engine
Mohamed Hammadh Hassan, Zailan Arabee Abdul Salam, Kadhar Batcha Nowshath · 2022 IEEE International Conference on Distributed Computing and Electrical Circuits and Electronics (ICDCECE) · 2022
Maldives is a small country located at the Arabian Sea of the Indian Ocean with a small population speaking and writing in their own language - Dhivehi. Although Dhivehi content has increased significantly over the years, it is still quite lacking in the digital literature tools domain. Therefore, most of these contents are often stored as physical copies, or digital scanned images. To ensure that these documents are not lost, it is important to digitize them. In order to eliminate the need of manual data entry as well as serve as a first step to building a storage repository for Dhivehi documents, a Dhivehi OCR engine using image processing techniques, a unique character level segmentation approach, and a character level CNN model is proposed.