vCity Map: Crowdsensing towards visible cities
Yoshito Tobe, Itaru Usami, Yusuke Kobana, Junji Takahashi, Guillaume F. Lopez, Niwat Thepvilojanapong · 2014
Crowdsensing is a powerful system to utilize the activities by people carrying mobile phones embedded many kinds of sensors. We have developed a system called vCity Map which uses crowdsensing to visualize the city environment for two kinds of information: sound and road conditions. For sound, we analyze the sound of passing automobiles. Collecting sound at many places help understand the atmosphere in the urban environment if the sound information is associated with time and place. For road conditions, bicycle riders who carry smartphones automatically upload the information. The acceleration signals obtained by the smartphones are separated into two signals, movement of the users' body and road condition by the combination of Independent Component Analysis (ICA) and Fourier Transform. Both of these two kinds of information is analyzed at a cloud side. Our work differs from the previous work in that analyzed information is mapped into the map and road condition is obtained using ICA.