• Vinod Kumar K P
• Kavya CL
• Aishwarya B
• Kruthika K
• Kruthik D

Visually impaired people often find it difficult to move confidently in busy or unfamiliar places because they cannot easily sense the objects around them. Tools like white canes and guide dogs help, but they cannot tell what an object is or whether it is moving. To improve this, we present a real-time system that can recognize and track objects using deep learning. It uses a camera along with YOLOv5 to detect objects and Deep SORT to follow their movement. The system also estimates how far the objects are and from which direction they are coming. All this information is then spoken to the user through an offline Text-to-Speech engine, giving clear and timely audio guidance. The system runs on portable devices like Raspberry Pi 4 and Jetson Nano, ensuring fast performance without needing internet. Tests in real environments showed high accuracy, smooth tracking, and quick voice alerts, proving that the system can meaningfully improve safety, awareness, and independent mobility for visually impaired individuals.

Object Recognition, Object Tracking, Deep Learning, YOLOv5, Deep SORT, Real-Time Detection, Text-to-Speech (TTS), Raspberry Pi.

"OBJECT DETECTION AND RECOGNITION TO ASSIST VISUALLY IMPAIRED PEOPLE", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 11, page no.f767-f771, November-2025, Available :http://www.jetir.org/papers/JETIR2511612.pdf

"OBJECT DETECTION AND RECOGNITION TO ASSIST VISUALLY IMPAIRED PEOPLE", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 11, page no. ppf767-f771, November-2025, Available at : http://www.jetir.org/papers/JETIR2511612.pdf