Detecting traffic congestions using cell phone accelerometers.

Mingqi Lv,Daqiang Zhang,Ling Chen,Gencai Chen

DOI: https://doi.org/10.1145/2638728.2638744

2014-01-01

Abstract:ABSTRACTIn this paper, we propose a system that detects traffic congestions by using cell phone accelerometers, which have many advantages (e.g. energy-efficient, unobtrusive, impervious to environmental noise, etc.). However, it is challenging to extract well-targeted and accurate features (e.g. speed) for detecting traffic congestions in a complex daily-living environment using a single cell phone accelerometer. The proposed system comprises a vehicular movement detection module, and a module for likelihood estimation of traffic congestions. Experimental results based on real datasets have demonstrated the effectiveness of the proposed system.