Background

Environmental　sounds have large influence on our daily life. In a city, for example, noise pollution has been a big problem, and noise often becomes the cause of trouble between the next-door neighbors. On the other hand, a sound of a stream, or a rustle of trees gives an extremely high relaxing effect to a person. To improve city life and to design a smart city, it is important to properly deal with the environment sounds.

Objectives and Expected effects

We have developed a sound collection system based on a crowdsourcing approach using smartphones as sensors, which results in collecting environmental sounds from many users with their locations as context information. Moreover, we developed an algorithm to estimate crowed situations using collected sound data. Because the algorithm runs on smartphones and only the crowed situations are sent to cloud servers, there is no risk to leak out sensitive information through the sounds. The long-term and short-term collection of the crowed situation are useful for a city design and an event control, respectively. We have also been investigating the method of visualizing crowded situations on maps in real-time manner. To confirm allover performance of the system, we carried out field tests in two Japanese cities, Okayama and Kurashiki.
 

URL

http://https://www.a.cs.okayama-u.ac.jp/scope_h27/