Recent developments in satellite sensors tend to the availability of high
spatial and spectral resolution images. The motivation of this research
paper is to get maximum benefits of different bands in high resolution
satellite images. In this research paper, a novel classification technique is
introduced where the shared texture features properties problem is
addressed. Atmospheric correction is applied on a high resolution World
View 2 (WV2) image to produce reflectance value for all spectral bands.
Reflectance image is produced by knowing the environmental parameters
of the images at the capturing time, which can be extracted from auxiliary
files associated with the input image. A multi-layer classification tree
analysis is applied on a reflectance image to extract urban area features
based on investigated thresholding values. The proposed technique is
investigated through MATLAB environment. The results of the proposed
technique are assessed versus classification results of Maximum
Likelihood classification technique that is applied through ENVI software.
The assessment of classification results is represented in confusion matrix
format and determination of Kappa Coefficient. The investigated
technique succeeded in classifying urban area features up to 90%. The
proposed technique is fast, automated and suitable for any image with
same spectral bands as WV2 satellite image.