Nepali Handwritten Character Recognition System
Santosh Acharya, Shashank Dhungel, Ashish Kumar Jha · Zenodo (CERN European Organization for Nuclear Research) · 2022
Even if the technological and digital world is expanding more quickly, there are still many things that are lacking. What a wonderful thing it would be to be able to trust machines to scan any handwritten characters into digital representation. The method for doing this is called optical character recognition (OCR), but there is still much room for improvement. Although there has been work done on it, the technique developed for one language cannot be applied to another due to language variations. Nepali is not a language that is frequently used online. Perhaps this is why there are fewer OCR systems developed using this language. We have made an effort to improve on it so that Nepali characters can be recognized. Basically, the idea is to use a camera to scan Nepali handwriting from hard copy paper, locate the regions in the image where the characters are present, segment those localized parts into characters, and then digitally display each predicted segmented character.