抄録
This paper describes methods of producing a built-up area map and a densely built-up area map by using TM (Thematic Mapper) data taken on November4, 1984 acquired from LANDSAT-5. Through a case study of Kanagawa Prefecture, this study tries to analize by overlaying and comparing LANDS AT MAP with DID (Densely Inhavited Districts) or CBD (Central Business Districts) which we devide cities with lines on the map. The land-cover is classified into nine categories; high-rise building area, middle-rise building area, low-rise building area, concrete pavement, bare land, lawn land, grass land, trees and water, using level slice method with training areas shown in Table-3. The regression analysis between TM data and 500 m mesh data in Tokyo shows that the grasp of the built-up area pattern, as well as natural surface pattern, is possible. The process of producing a built-up area map is shown as follows : 1) land-cover classification 2) make up 100m mesh (4×4=16 pixels) 3) decide the constituent element of the built-up area map (high-rise bldg. Area, middle-rise bldg. Area and low-rise bldg. Area) 4) make up 100m mesh data of the built-up area (Fig.6 and Fig.7) 5) map output by personal computer The process of producing a densely built-up area map is shown as follows: 1) land-cover classification 2) make up 50m mesh (2×2=4 pixels) 3) decide the constituent element of the densely built-up area map (high-rise bldg. Area and middle-rise bldg. Area) 4) make up 100m mesh data of the densely built-up area (Fig.10) 5) map output by personal computer Concretely, we produce a built-up area map of Kanagawa Prefecture, shown in Fig. 9, and densely built-up area maps of Yokohama City, Odawara City, Yokosuka City, Kawasaki City and Fujisawa City, shown in Fig.11〜Fig.15. The results show that possibility and effectiveness using remote sensing as one of the national land informations on arban area.
