SMART VISION AID: INTEGRATED AI SYSTEM FOR OBSTACLE DETECTION, FACE RECOGNITION, CURRENCY IDENTIFICATION FOR THE VISUALLY IMPAIRED

Abstract

Assistive technologies for visually impaired individuals have evolved significantly with the advancement of Artificial Intelligence and embedded systems. However, many existing systems focus on a single functionality such as obstacle detection or currency recognition. This paper proposes an AI-based Smart Assistive Vision System implemented using Raspberry Pi 4 and computer vision techniques. The system integrates obstacle detection, face recognition, and currency identification into a unified portable device. A camera module captures real-time visual input, which is processed using OpenCV and deep learning models. The recognized information is converted into audio feedback using a Text-toSpeech (TTS) engine. The system ensures real-time processing, cost-effectiveness, and portability. Experimental results demonstrate high detection accuracy and reliable performance under various environmental conditions. This integrated assistive system enhances independence, safety, and social interaction for visually impaired user.

Keywords