Assistive Glasses for the Visually Impaired Using Raspberry Pi with OCR, Object Detection, and Speech Output
Year: 2026
Type: conference-paper
Abstract: This paper presents the design and development of assistive glasses for visually impaired individuals using a Raspberry Pi 4 Model B as the core processing unit. The proposed system integrates YOLOv10 for real-time object detection, Tesseract OCR (Optical Character Recognition) for printed text recognition, and the eSpeak text-to-speech engine for audio feedback, housed in a compact wearable eyeglass frame. Unlike existing assistive wearables that address OCR or object detection in isolation, this system combines both functionalities in a single, cost-effective, offline-capable device. A physical mode selection button allows the user to toggle between OCR mode and Object Detection mode. The system is evaluated through indoor and outdoor testing across multiple lighting conditions, measuring detection accuracy, OCR accuracy, and end-to-end response time. Performance benchmarks are compared against prior systems using YOLOv5 and YOLOv3, with the YOLOv10-based prototype targeting superior accuracy and faster inference on resourceconstrained hardware.
Institution: Mapúa University
Language: English
Subfield: Occupational Therapy
Field: Health Professions
Domain: Health Sciences
Open Access status: hybrid
OpenAlex ID: W7204169031