ODTÜ Bilgisayar Mühendisliği Bölümü CEng 564 Dersi 2010 Güz Dönemi Ders Projei
Satellite images become abundant; as a result commercial applications like GPS devices, Maps and City Planning are more accessible. Detection of building in City planning is very important to analyze growth of a city. Satellites provide data in different wavelengths. Apart from Red, Green and Blue bands, satellites usually provide data in Near Infrared and other spectral bands. In this paper, Building detection using multi-spectral bands investigated. Segmented images have been classified using k-NN classifier. Effect of Pre-processing, first, second and third moments, and HSV color space also evaluated.
