OptaliX: A Wearable AI-Based Vision Assistance System for Visually Impaired Users

OptaliX is a wearable, AI-based vision-assistance prototype designed to help visually impaired individuals access visual information such as objects, currency, printed text, products, and their surrounding environment through real-time audio feedback. Built on a Raspberry Pi Zero 2 W paired with an Arducam 8 MP IMX219 ultra-wide-angle camera, the system integrates object detection, currency recognition, product recognition, expiry-date recognition, and scene analysis into a single modular perception pipeline. Experimental testing showed approximately 91% accuracy in object detection with negligible latency, 82.69% accuracy in currency recognition (43/52) with 3–4 second latency, 100% accuracy in expiry-date recognition (55/55) with 5–6 second latency, and 92.73% accuracy in product recognition (51/55) at similar latency, while scene analysis produced correct interpretations across all evaluated inputs at approximately 7-second latency. These results demonstrate the technical feasibility of combining multiple visual-assistance functions on a single low-cost wearable platform. This work is presented as an initial feasibility study and does not claim clinical validation or readiness for large-scale deployment.