Effect of Atmosphere Correction for Unsupervised Classification in Multispectral Camera LAPAN A3

On their way to the Earth’s surface, electromagnetic waves emitted by the Sun are subject to atmospheric distortion in the form of refraction, scattering, and absorption events. As a remote sensing satellite, LAPAN A3 imagery is not immune to atmospheric disturbance. To produce images free of atmospheric disturbance, corrections are needed to remove atmospheric effects. In this study, a comparison was made between three atmospheric corrections, namely histogram adjustment, FLAASH and 6S. The research was conducted in the Batu Jamus rubber plantation area owned by PT Perkebunan Nusantara I Regional 3 and the Karanganyar area. Atmospherically corrected images were used for unsupervised classification using the ISODATA method. Visual analysis showed that atmospheric correction of the LAPAN A3 image resulted in a brighter image in the FLAASH method. Spectral reflectance curve analysis shows that the histogram adjustment method has a similar spectral reflectance shape to the reference. The classification results show no change in the histogram adjustment method, which is indicated by the accuracy value reaching 100%. On the other hand, for the FLAASH and 6S methods, there is a change in the classification, indicated by a decrease in accuracy of 44.0514% and 48.8370%, respectively. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.