Hybrid machine learning in classification methods for HCR in gujarati language

Abstract

The problem of recognizing Gujarati Handwritten character with vowels opening new future scope where one can use smart phone, website or any handy scanner to convert hand written Gujarati Language into text. It will be very effective to give education in mother language at primary level. Public, Private and Government sectors will be benefited when they get any hand written Gujarati Script and they can directly convert it into softcopy or into text form. There are many methods used to solve this problem.Using CNN we can improve new algorithm depending on training data set, mathematical model and other intricacy. Convolutional Neural Network or machine learning is very useful for this. Still there are more chances for improvement and rising accuracy using Machine learning in combination of Deep Learning as a hybrid model.