PENGARUH KOMBINASI TRANSFORMASI SPEKTRAL TERHADAP HASIL KLASIFIKASI KEPADATAN LAHAN TERBANGUN

The utilization of remote sensing imagery with medium resolution for mapping to identify the built-up area density is not optimize yet. This study aims to assess spectral transformation and combination transformation ability for detection built-up area density. This is an experimental research by developing the spectral transformations to find out the built-up density pattern from the best spectral transformation result. Transformations used are NDVI, UI and the combined transformations used are UI-NDVI. The best transformation single is UI transformation with accuracy point about 83,33%; kappa coefficient agreement is ( Ќ) 0.78; and deviation standard is 30.27. According to kappa coefficient this result is categorized as medium confidence level. While the best combined transformation is UI - NDVI transformation with accuracy point about 85.04%; Kappa coefficient agreement (Ќ) 0.82; and deviation standard is 30,32. This combined transformation reaches the high confidence level. This research based on combined spectral transformation for built-up area dencity mapping proven to be increase result classification accuracy.

1.  UEU-Journal-20231-11_1374.pdf - 1275 KB