Yoshihiro Kosaka

Learn More
This paper explores a method for creating large-scale urban 3D models using Historical GIS data. The method is capable of automatically generating realistic VR models based on GIS data at a low cost. 3D models of houses are created from polygon data, fences from line data, and pedestrians and trees from point data. The method is applied to the Virtual Kyoto(More)
Recent advance in ubiquitous technology has led the development of ubiquitous sensor environment in which various kinds of sensor devices provide sensor data. In this environment, since an enormous number of sensor devices join the network, it is effective to construct a P2P network for efficient retrieval of sensor data. In general, each sensor device(More)
In this paper, we propose a P2P network construction method to achieve effective sensor data collection in a ubiquitous sensor environment. Our method constructs a P2P network overlay considering the location and the cycle of sensing of each sensor device (peer). First, according to the location of each peer, a hierarchical overlay network is constructed.(More)
This paper explores a method for creating large-scale urban 3D models from GIS spatial data. It is capable of automatic generation of realistic VR models based on existing GIS data. Parametric 3D models of houses are created from polygon data, fences from line data, and pedestrians and trees from point data. The method is applied to the project(More)
This paper explores a method for reconstructing large-scale 3D urban areas from 2D data of the Geographic Information System (GIS). Our approach enables automatic reconstruction of 3D models of houses using the GIS data. The main target of the automatic generation of 3D models is the city of Kyoto in the Edo era and the houses peculiar to the urban area of(More)
  • 1