Improving the safety of cyclists is a critical traffic issue. Cyclists’ riding speed and acceleration often indicate their intentions. Currently, most autonomous vehicle manufacturers are working toward developing vehicle automation technologies that fully rely on the vehicle’s onboard sensing and intelligence. However, onboard vehicle intelligence may still encounter difficulties when detection capability or range is physically obstructed or weakened. This paper proposes to develop a speed estimation function intended to be applied in cyclists’ smartphones. The proposed system provides the accurate speed of cyclists in an effort of establishing vehicle-to-everything communication. By making use of the accelerometer and gyroscope sensors in a smartphone attached to a bicycle, cyclists’ speed can be estimated using one of the most prominent deep neural networks, namely Long Short-Term Memory. The experiment results demonstrate the proposed method can outperform the conventional speed estimation method.
Cyclist Speed Estimation Using Accelerometer and Gyroscope in Smartphones
2022-03-07
1094802 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Europäisches Patentamt | 2022
|Speed Estimation using Smartphone Accelerometer Data
Transportation Research Record | 2019
|