Comparison of Pixel and Object Based Approaches Using Landsat Data for Land Use and Land Cover Classification in Coastal Zone of Medan, Sumatera

Abstract

As an archipelagic country, Indonesia has the second longest coastal areas in the world after Canada. Coastal zone has dynamic characteristics. There are many changes here because of an unique ecology systems, sedimentation as well as human activities. Because of the coastal zone dynamics especially land use and land cover therefore it is important to identify them by using remote sensing technology. This paper discuss the application of Landsat satellite remote sensing image to classify land use and land cover by using object based classification approach in part of coastal zone of Medan, North Sumatera, Indonesia. Conventional classification methods use per pixel approaches that rely only on the spectral information or colours contained in the image. Otherwise object based classification approach firstly the image is segmented into objects. In subsequent steps, segments are merged based on their level of similarity. The user uses a scale parameter which indirectly controls the size of objects by specifying how much heterogeneity is allowed within each. User-defined color and shape parameters can also be set to change the relative weighting of reflectance and shape in defining segments. The methodology is consisted of satellite data acquisition, existing topographic map and statistical data collection, rectification of Landsat image, classification of land use and land cover using maximum likelihood algorithm and object based approach. Finally, the result shows that use of object based classification system provides reliable classification result than using traditional method such maximum likelihood classification system.