BikeSafe
Weixi Gu, Yunxin Liu, Yuxun Zhou, Zimu Zhou, Costas J. Spanos, Lin Zhang · 2017
Monitoring the bicycle safety is of great importance. The current methods either require specific hardware supports or are expensive to implement. In this paper, we propose BikeSafe, a smartphone-based system to track bicyclist movements and alarm their dangerous riding behaviors in real time. Preliminary experiments over 12 participants show that the overall detection accuracy of BikeSafe on riding behavior achieves 86.8%, and that of the illegal way riding reaches around 90%, satisfying the practical operation in daily usage.