Department of Computer Science & Engg. Rungta College of Engg. & Technology Bhilai, Chhattisgarh, India

Abstract

Extraction of objects from high spatial resolution images is one of the most desired and exciting topics in research area of surveying and mapping. The previous work has been done for clipping same type of objects and some act has been done for improving its efficiency. Level set evolution is used to give effective outcomes for clipping objects more efficiently from high spatial resolution images including physio-graphical changes. In previous paper level set evolution (LSE) is used for taking out unnatural objects. Two types of LSEs are used i.e. edge based LSE and region based LSE. Some approaches are used for change detection from high resolution remotely sensed images. An active contour model helps to detect building rooftops and building boundaries. In comparison of other methods LSEs are more remarkable and accurate for clipping out unnatural or man-made objects.

Keywords: Extraction of Natural Objects, Extraction of Unnatural Objects, Level Set Evolution (LSE), High Spatial Resolution Images

View PDF