Sound power level determination of roads in Tokyo using aerial photographs, machine learning, and ASJ RTN-Model 2018

Abstract

The traffic noise map is a visual representation of the sound level distribution in and around a road based on the sound power levels of vehicles on the road and the diffraction and reflection by the surrounding buildings. The purpose of creating the traffic noise map is to facilitate the government, researchers, industry, and the public in their efforts to track trends in transportation-related noise. To create noise maps, sound power level determination is the most important. The road traffic noise prediction model "ASJ RTN-Model 2018" has been widely used for road traffic noise prediction in Japan. Provided a method of calculating sound power level by vehicle classification, traveling speed, and driving mode. In the study, to obtain the vehicles' classification and coordinates, YOLO (You Only Look Once) was used to process the aerial photos. This study constructed sound power level distribution on the road network in Tokyo. Additionally, the estimated data were compared with the measured data to confirm the estimated data's reliability.