In this study, Landsat 5-TM data were used to map urban land classes and the changes that occurred within them over a period of six years. The land classes were identi® ed by Landsat 5-TM scenes taken in the same season in 1988 and 1994. The phenomena of land class changes were evaluated by adopting two remote sensing approaches, namely mapping and modelling, in a case study of the Bangkok Metropolitan area of Thailand. The quantitative results of changes, which were computed from a post-classi® cation method, were used to analyse the pattern of changes in the urban land classes. The change-detection analysis indicated that 2% of agricultural land was lost, and there was a 14% increase in the commercial areas. The results demonstrated that the pattern of change in the urban land classes in Bangkok was that of agriculture lands to open lands; open lands to residential, and residential to commercial. The highest commercial land growth was observed in the high-density residential areas along main roads and the railway line.Data were generated from the two dates of TM images for the vegetationimpervious-soil (V-I-S) composition model. The trends of changes in the urban land classes and the anatomy of the study area were presented quantitatively through the V-I-S model. Good agreement was obtained between the values of changes computed for the impervious surfaces from the V-I-S model (which showed 6% changes) and the change-detection map (which showed 5.6% changes). The results of changes in the spatial pattern of commercial and residential areas ( high, medium and low) emphasize that remote sensing data can be used for V-I-S modelling and mapping of urban surface features.

No abstract

A methodology has been formulated to integrate images from IRS-1A LISS II of two dates for landuse/landcover classi cation. The methodology developed includes image classi cation by fuzzy k-means clustering and fusion of memberships by fuzzy set theoretic operators. The two date images have been geometrically coregistered and classi ed for the identi cation of land classes individually. The fuzzy memberships of the classi ed output images have been integrated by using fuzzy logic operators like algebraic sum and gamma (c) operator. The classi cation accuracy of the resultant land classes in the integrated images was veri ed with the ground data collected in situ. The resultant images have been evaluated by kappa (k) statistic and it was found that output from the image of fuzzy algebraic sum operator scored high in generating the land classes, with an overall accuracy of 95%.

No abstract

No abstract