Advances in Handwritten Character Recognition: A Comparison of OCR and Large Language Model-Based Approaches

Handwritten documents are used daily for communication and information restoration. The process of identifying handwritten characters from pictures and PDF documents and transforming them into a form that machines can read for additional processing is known as handwritten character recognition. Because handwritten documents have distinct features such as marginal comments, overlapping letters and phrases, and varying writing styles, digitizing them can be difficult. In this research, We present two different models Large Language Model (LLM) and Optical Character Recognition (OCR) for handwritten character recognition. With a wide range of pattern recognition capabilities, The algorithm for optical character recognition is specifically utilized for the extraction and identification of characters. Large Language Model operations have been used to extract text from images. This work's primary goal is to examine both OCR and LLM's capacity to identify characters from a picture collection and suggest the best method to improve accuracy and efficiency in solving handwriting recognition challenges.