The effect of atmospheric correction on object based image classification using SPOT-7 imagery: a case study in the Harapan and Kelapa Islands

Abstract

Object-based classification (OBIA) has become a new paradigm in remote sensing data analysis. Errors in interpretation of benthic habitat classification image can be caused by deviation of values in reflectance. The absorption, scattering, and atmospheric reflections caused differences between the reflectance value of satellite images and real objects. The author have tried to compare two different approaches of OBIA classification with FLAASH atmospheric correction and OBIA classification without atmospheric correction. In this research, atmospheric correction with FLAASH method on SPOT-7 imagery conducted to eliminate atmospheric effect on the image. The Support Vactor Machines (SVM) algorithm is selected for the classification of benthic habitat in Harapan and Kelapa Island. Atmospheric-corrected image was analyzed visually then compared to ortho image in reflectance pattern, object segmentation and classification result. Results showed an improvement of visualization and reflectance pattern in shallow water objects. This study concluded that the FLAASH method as one of the appropriate methods to be used in atmospheric correction in SPOT-7 images.