Real-Time Object Recognition with Voice Feedback for Visually Impaired Using Raspberry Pi

This paper presents a real-time object recognition system with voice feedback designed for visually impaired individuals using a Raspberry Pi platform. The system integrates computer vision, deep learning, Optical Character Recognition (OCR), and sensor-based technologies to provide environmental awareness and safety. A USB camera captures real-time images, which are processed using a YOLO-based object detection model. Detected objects are converted into speech output through a text-to-speech module. Additionally, an OCR module enables users to read printed text aloud. An ultrasonic sensor is used for obstacle detection, providing audio warnings when objects are nearby. The system also includes an emergency alert feature that sends the user’s location via email using a GPS module. The proposed system is compact, cost-effective, and energy-efficient, making it suitable for portable assistive applications. Experimental results demonstrate that the system effectively improves independence, safety, and accessibility for visually impaired users.