Published Paper Details:

In the current digital era, technology has become omnipresent in people's lives and some human tasks such as face recognition, handwritten recognition, voice recognition are now gadget tasks. Pattern recognition has a wide range of applications, one of which is handwriting recognition. Handwritten recognition allows a machine to detect the characters written in a user's handwriting automatically. It is concerned with extracting the majority of the necessary information from raw data. Deep learning is perhaps the most sophisticated machine learning approach widely available.Offline Handwritten Recognition, which works on scanned images, and Online Handwritten Recognition, which requires input at run time. Offline handwritten handwriting is the most tricky to recognize. Hence, the present idea is to implement a Real-Time Hand-Written Character Recognition System using a Convolutional Neural Network(CNN). The suggested system will extract texture features to detect handwritten letters, words, or text. The flexibility of establishing this intended character recognition system by integrating it into an existing system has been enhanced. This system will be divided into three parts: training, testing, and recognition. CNN is used as the training algorithm and then tested on the IAM words and lines dataset. The IAM dataset has a total of 96000 handwritten images of words and a total of 13353 handwritten images of lines. The proposed model classifies with 89 percent overall accuracy.