Deep Learning Based Gujarati Handwritten Character Recognition
Dhara S. Joshi, Yogesh R. Risodkar · 2018 International Conference On Advances in Communication and Computing Technology (ICACCT) · 2018
Optical character recognition is a method of the renovation of pictures of printed text, handwritten documents andtypewritten documents into a document recognized by the machine. This can be used for editing, reduction in storage space etc. So, basically, it relates to posses of teaching and recognition of data. Hence, here a hand written document which is in a form of some local language characters except than English. In this paper, we are using the Gujarati Characters for the recognition into the machine editable format. Deep learning is the method used for identity recognition which will depend on the man-made neural network. Deep learning involves a proper focus on right features of the image to recognize themselves with a negligible assistance from the program. Neural network evolved considering the concept of biological neurons. Accuracy and efficiency are improved by use of deep learning for character recognition.