Change Detection from Multitemporal Land-cover Class Images

Earth-observing satellites, such Landsat, provide many multitemporal images of earth and with classification techniques, land-cover class images can easily be prepared. Multitemporal images mean several images which were recorded/acquired at difference times. Whenever there are two or more multitemporal land-class images, two questions commonly raised are: ‘where are the areas changed’ and ‘change to what.’ The pattern of change analysis technique in this study is trying to provide quick information which is needed to answer those questions. In this study, 9 Landsat images with different acquired times were used and classified into 3 land-cover classes in order to give a simple understanding on how the technique worked. Each Landsat image is in raster format, and with 9 images available as time series, those multitemporal images formed a cubic raster database and the information will be retrieved from this one.

• Type: Conference
• Year: 2006
• Where: MOMM
• Authors: Petrus Paryono
• Keywords: Images, Landsat Image, MOMM 2006, Multimedia, Multitemporal Images