부산대학교 도서관

This paper presents a geospatial analysis of a FlatGeobuf database composed of six years of geographic data on approximately 41.7 million Mapillary photo shooting locations throughout Japan and geospatial data including road data from OpenStreetMap (OSM). Although Mapillary has a shorter track record than OSM, it is a massive data source and its use as a new resource for volunteered geographic information is expected to attract attention in the future. Therefore, we aim to clarify the geographical distribution of Mapillary users and time-series transitions and attempt to analyze the relationships of these road landscape images with OSM road data and the local geographical distribution characteristics. Although the geographical distribution of Mapillary locations is biased, much information is collected for local roads, which may improve the quality of OSM road data and expand the overall road data.