Pan-Sharpening Performance Comparison for Land Use Classification Application, and Its Effect on LISA LAPAN-A3 in Accuracy Improvement

Pan-sharpening is one data fusion application that aims to increase the spatial resolution of the multi-spectral image by merging a low-resolution multispectral image with a high-resolution panchromatic image. This process is commonly used to increase the quality of images in the application of land use classification. This research aims to see and learn about the performance of the pan-sharpening method in terms of Land Use Classifications. 5 different methods are compared to see each performance in classification. Moreover, not only using a single-platform data, which is multi-spectral (MS) and panchromatic (Pan) image from Landsat 8, this research also tries to fuse 2 data from a different platform, which are MS from LISA LAPAN-A3 and Pan from Landsat 8. It found that each pan-sharpening method has a different result in terms of accuracy when applied to single-platform data and cross-platform data, nevertheless, some improvements in accuracy were slightly found in pan-sharpened LISA's product to a 9.31% increase.