Abstract
Reconstruction of lost historical city or spot has recently received more attention from a possibility that people can appreciate or experience the historical space through the computer. In general, great deal of time, labor and skill are needed to reconstruct 3D historical objects. In particular, 3D modeling of buildings are important issue since buildings are an important component of historical city model.
This paper describes an efficient city modeling method for visualization of historical city, and visualization of historical spot, around Hakone Barrier Station is demonstrated.
1. Introduction
Existing colorful historical maps were mostly produced at the Edo period in Japan, and these historical maps which show situation of land use and streets in those days often give important information for studying history of city planning, civil engineering, architecture and so on. In particular, visualization of historical city helps to make understanding for situation at the time when the maps were produced more efficiently. On the other hands, virtual archeology has recently received more attention from possibility that people can appreciate or experience the archeological objects or historical space and art through the computer at any time and without going to the museum.
The efficient method for visualization of historical city using the historical maps was proposed by the authors (Namatame & Chikatsu, 1999), and the remarkable points of this method were its ability to arrange each building on the historical map automatically.
Hakone locates near Mt. Fuji, and Hakone barrier station was the most important station for the Edo period since the Japan was divided into two territories at the time. As Hakone barrier station was lost, only the sits remain, Hakone Barrier Station is reconstructed from the view point of digital archive and VR museum, and visual simulation around the Hakone Barrier Station is demonstrated in this paper.
2. Efficient City Modeling
2.1 Extraction of Dwelling Area
Figure 1 shows historical map used in this paper. The historical map is around the Hakone Barrier Station, and the circumference for each attribute data such as inns is drawn by black line in this map. The extraction method of the attribute data using mathematical morphology and remapping method were described by the authors (Yamada & Chikatsu 1999). However, extraction for small characters included dwelling area and insufficient clarity color were issues. Then, in order to acquire information for dwelling area such as position, area, height and width ratio, kind of dwelling area, firstly black lines in figure 1 are extracted using Gauss-Laplacian(GL) filter and extracted lines are corrected by manual. Figure 2 shows the corrected edge image. Secondly, position, area, height and width ratio for dwelling area are acquired by labeling procedure.
