Use of multispectral data for vegetation cover identification through subpixel classification technique. | [Journal of Scientific Research • 2006]

Author(s):

1. Khunsa Fatima: Department of Space Science, University of the Punjab, Lahore, Pakistan

2. Asim Daud: Department of Space Science, University of the Punjab, Lahore, Pakistan

3. Urooj Saeed: GIS Lab, WWF-Pakistan, Lahore, Pakistan

4. Salman Ashraf: GIS Lab, WWF-Pakistan, Lahore, Pakistan

Abstract:

Over the last decade, traditional classification techniques have been used to extract landcover information from remotely sensed images and were limited to the pixel level only. However, single DN values of satellite image pixels are often comprised of reflectance responses from different landcover features. This mixed pixel problem is the actual hurdle in accurate landcover classification which is significant in medium to low resolution remotely sensed data. In the present research study, subpixel classification technique is used to address the mixed pixel problem in medium resolution Landsat satellite data where as high resolution data was used for the selection of training areas and for accuracy assessment. The results showed that subpixel classification technique is time saving in comparison to traditional classification techniques.

Page(s): 53-58

DOI: DOI not available

Published: Journal: Journal of Scientific Research, Volume: 35, Issue: 1, Year: 2006

Keywords:

Keywords are not available for this article.

References:

References are not available for this document.

Citations

Citations are not available for this document.

Citations

Downloads

Views