Object-based Classification of Buildings from High Spatial Resolution Remotely Sensed Imagery Using Imagine Objective "case study of residential area in Damascus"

Abstract

Objects extraction from satellite images is one of the most important applications of remote sensing. The extraction of spatial information from remote sensing scenes is often carried out by laborious and costly manual digitization techniques. Despite scientific efforts to develop methodologies for classifying remote sensing scenes, most of these methodologies are still far from practical use. Most of the methodologies used for remotely sensed data classification are divided into two main domains: the first is pixel-based and the second is object-based classification. This research presents the common methodology available in the Imagine Objective program, which provides group of general tools that can be adapted to the users' needs and the specific objects that are to be extracted from the images.