Eco-Eye: Object Detection System for Blind People

Authors- Assistant Professor Deepali Mane, Rahul Kolhe, Mohit Patil, Vaishnavi Bharambe, Aarya Raghuvanshi

Abstract-This paper presents a research study on a smart walking stick designed to enhance mobility for visually impaired individuals. The system uses a Raspberry Pi with a camera module to process live video via the YOLOv3 object detection algorithm trained on the COCO dataset. It detects obstacles and hazards, providing real-time audio feedback through a speaker. The study details the hardware and software design, surveys related assistive technologies, and explores advancements in computer vision and deep learning. Edge detection, sensor fusion, and embedded optimizations ensure portability, affordability, and reliability. Experimental results show improved navigation safety, positioning this as a viable low-cost assistive solution. Future enhancements include multi-sensor integration, voice interaction, and cloud- based processing for next-generation assistive technology. Index Terms – Smart Stick, Object Detection, YOLOv3, Raspberry Pi, Assistive Technology, Blind Assistance, COCO Dataset, Edge Detection, Real-Time Processing.

DOI: /10.61463/ijset.vol.13.issue1.158