Abstract: Handwritten recognition has been one of the most challenging researches. . Major motives consist of the styles, and strokes of the huge variety of handwritings. Handwritten recognition is the cap potential of a pc to acquire and interpret handwritten entries from sources consisting of paper documents, photographs, contact screens, and different devices. Present techniques in the field of Handwritten Text Recognition are Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Ngram, and Long-short Term Memory (LSTM). They predict the handwritten word with good accuracy.
Language translation cannot always translate a language 100% because of the slang structure and linguistics. Machine translation techniques are majorly used for language translation to get culturally and linguistically appropriate translations. In our project, we have used CNN and LSTM (BLSTM) for handwritten recognition and For Language recognition, we have used Encoder-Decoder using LSTM. Our main objective is to combine handwritten recognition and language translation to interpret handwritten words appropriately and translate them into one of the native languages in India (Hindi) acquiring good accuracy.

Keywords: Handwritten recognition, IAM dataset, Language Translation

PDF | DOI: 10.17148/IJARCCE.2022.115202

Open chat
Chat with IJARCCE