Author/Authors :
Saleh, Salah A. H. ministry of science and technology, Iraq , AL-Kabi, Enaam K. H ministry of science and technology, Iraq , Karm, Nehad A. University of Baghdad - College of Science - Department of Astronomy, Iraq
Abstract :
in this research digital numbers (DNs) for Landsat-5 TM images was Converted to reflectance values, at first DNs converted to radiance by corrected for Gain and Offset values of the particular spectral TM band. Then converted these radiance values to spectral reflectance by correcting the atmospheric effects To study the effects or the conversion, digital image processing techniques were applied for both images (DNs and Reflectance). such as enhancement by histogram equalization, and geometrical correction. Image classification (supervised and unsupervised) also applied as comparative criteria. The analysis of the results shows that the reflectance image gave better classificaion accuracy for supervised and unsupervised classification than image DNs while conversion DNs to reflectance image does not affect ns geometrical classification than image DNs while conversion DNs to reflectance image does not affect ns geometrical .