Real-Time LiDAR-Based Stair and Elevation Change Detection on Consumer Mobile Hardware: A Single-User Study
This paper presents and evaluates a computationally efficient, real-time algorithm running on consumer LiDAR-equipped smartphones that achieves a 92.5% detection rate for stairs and elevation changes with zero false positives, demonstrating the viability of low-cost, on-device assistive navigation tools for the blind and visually impaired despite inherent geometric limitations in descending approaches.