Date of Award
6-3-2026
Document Type
Thesis
Publisher
Santa Clara : Santa Clara University, 2026
Department
Computer Science and Engineering
First Advisor
Krishna Ramamoorthy
Abstract
This project presents a wearable assistive system designed to help blind or visually impaired people navigate their surroundings more safely. The system uses a Raspberry Pi 5 connected to Camera Module 3 (NoIR Wide) sensors to capture a wide field of view. Real-time object detection is performed using a lightweight YOLOv8 model, and a redundancy-based merging algorithm combines detections across cameras to improve reliability. By requiring agreement between multiple cameras, the system reduces false positives and increases confidence in identifying obstacles.
The device is designed to operate in real time and provide reliable feedback about nearby hazards, helping users make safer navigation decisions. The multi-camera approach allows for better environmental awareness compared to single-camera systems, while the optimized software pipeline ensures efficient performance on embedded hardware. Overall, this project demonstrates a practical and scalable approach to assistive technology, with the potential to improve independence and safety for visually impaired users in everyday environments.
Recommended Citation
Daly, Nolan; Diaz, Sergio; and Ng, Brandon, "Wearable Hazard Detection System For the Visually Impaired" (2026). Computer Science and Engineering Senior Theses. 376.
https://scholarcommons.scu.edu/cseng_senior/376
