Viahero: gps-enabled radar smart cane integrating yolo-based obstacle and public utility vehicle recognition for enhanced mobility of visually impaired commuters/
Ma. Charina O. Gavica, Ma. Cristina F. Jara, Hannah R. Jovellanos, Dixie Marie Trista S. Marquez, and Khryzelle Trisha L. Suyat.--
- Manila: Technological University of the Philippines, 2025.
- xii, 214pages: 29cm.
Bachelor's thesis
College Of Engineering.--
Includes bibliographic references and index.
The problem of heavy traffic in urban areas in the Philippines is especially problematic for visually impaired individuals, given the limitations of white canes, which can only be used for close-range ascertaining of spatial obstacles. This study addresses the gap by proposing a GPS-enabled radar smart cane with YOLOv8 object detection on Jetson Nano, enabling autonomy and enhanced situational awareness in outdoor settings. The system provides tracking in motion in real-time with Neo-6M GPS for location, obstacle detection with RCWL-0516 radar, and detection and classification of stationary and moving public utility vehicles (PUVs) using a custom-trained YOLOv8 model. To further improve usability, the smart cane is designed with dual feedback mechanisms; audio alerts for object detection are spoken through wireless earphones, while haptic feedback from a vibrating handle gives cues for physical obstacles identified by the radar sensor. Training results of the YOLOv8 model showed a 92.9% mAP@0.5 detection accuracy. The runtime was also estimated when using a 20000mAh power bank to be 5.73 hours. In The User Evaluation with 13 visually impaired participants, average ratings were high: 4.4 for Functionality and Efficiency, 4.21 for Usability and Aesthetics, 4.23 for Mobility and Maintainability, and 4.29 for Safety and Reliability. This strongly suggests that the smart cane is very reliable in providing real-time alerts and navigation support greatly enhancing the user’s safety, mobility, and independence.