An IoT-Edge Smart Glass Prototype for Object Identification Assistance Using YOLO26

Abstract

Visual impairment can limit independent mobility, especially indoors where object-level awareness is needed beyond simple obstacle detection. This paper presents a low-cost IoT-edge smart-glass prototype for object identification assistance. The system integrates an ESP32-CAM for image acquisition, an Arduino Mega for sensor-feedback coordination, an HC-SR04 ultrasonic sensor for distance measurement, and a buzzer for auditory feedback. YOLO26 inference is offloaded to an edge-processing layer, while the embedded hardware handles sensing and feedback control. Preliminary indoor testing with low-quality camera frames showed that the prototype could identify common objects such as chairs and persons. The results demonstrate the feasibility of combining low-cost IoT hardware with edge-based object detection, although broader evaluation is required before claiming robust real-world performance.