Finger Reader for Visually Impaired People Using Optical Character Recognition (OCR) and Machine Learning Technique

Abstract Finger Reader is a wearable gadget that uses machine learning algorithms and a camera positioned on the finger to read text for those who are blind or visually impaired. The device takes text pictures and recognizes letters, words, and phrases using convolutional neural networks (CNNs). The World Health Organization estimates that 253 million people globally are blind or visually impaired. From the perspective of the visually impaired, there are a number of inconveniences that slowdown their day-to-day activities. To enable them to live in the modern world without difficulty, it is imperative that significant steps be taken using the technology currently in use. The system utilizes a combination of optical character recognition (OCR) and natural language processing (NLP) techniques to provide accurate and context-aware text recognition. The Device also includes a tactile feedback system, providing users with a physical representation of the text. Experimental results demonstrate high accuracy rates in text recognition, making Finger Reader a promising assistive technology for visually impaired individuals to access written information independently. Keywords: Optical character Recognition (OCR), Natural Language Processing (NLP), Convolutional neural network(CNN), Machine learning(ML), Finger Reader.