A CNN-Based Handwritten English Character Recognition

Recognition of handwritten documents allows data management and accessibility to be made easier in areas like archiving, finance, and health care. This research focuses on using CNN s to recognize handwritten English characters and words. The study uses a dataset of handwritten images for training and evaluation. The framework aims to address challenges related to variations in different writers' handwriting by leveraging CNNs for identifying handwritten English characters and words. The research achieved 94.36% accuracy in recognizing uppercase handwritten English characters and 88.97% accuracy in recog-nizing lowercase handwritten English characters. Additionally, 92.36% of uppercase words and 98.48% of lowercase words were recognized accurately.