MADVI

Visual impairment affects more than 2.2 billion people worldwide, with many experiencing limitations that reduce independence, productivity, and quality of life. Refreshable braille displays and other forms of assistive technology are vital to the visually impaired community to assist them in their daily lives. While technologies like these exist, current solutions on the market fall short of meeting several critical needs. Challenges include high retail cost, lack of physical-text recognition, limited real-time translation speed, inaccurate translation, concerns about user privacy, and a lack of portability for use in different environments. The MADVI project aims to directly address these gaps by developing a device capable of capturing printed text using a high-resolution camera, performing real-time optical character recognition through onboard artificial intelligence on a FPGA, and presenting the output through a fabricated refreshable braille display. This device keeps the privacy of users a priority by providing this functionality without relying on external connectivity or cloud-based services. The development of the device allows modularity for future innovation by using a serial bus to communicate between devices. The goal of this device is to cost significantly less than existing commercial products, possess a simple and intuitive interface suitable for both independent use and training environments, ensure strong privacy protection for sensitive documents, and provide a compact form factor comparable to leading portable braille displays. The MADVI system aims to provide a practical and empowering step towards equitable access to information for visually impaired individuals in a long-standing accessibility challenge.