Pemetaan Habitat Bentik Pulau Salarangan Menggunakan Metode Object-Based Image Analysis

Abstract

Salarangan Island has supporting ecosystems that are seagrass and coral reefs with a high increase in biological competence. Basic information about the distribution map of benthic habitats on Salarangan Island is not yet available so that technology is needed that can provide information spatially and temporally. Citra Sentinel-2A with Lyzenga transformation in mapping benthic and shallow habitats and studying the distribution of benthic habitats on Salarangan Island, Madura. The image data used is obtained free of charge from the website earthexplorer.usgs.gov and in situ data are used for image data validation. The multiscale / OBIA classification method with the SVM algorithm is used to classify objects in the navigation column. The classification results show as many as 4 benthic habitat classes with object / OBIA classification resulting in overall certification of 69.86%. Based on the classification results, the most contrasting class is the seagrass class (48.76 ha) because it has the most extensive area than the coral class (27.31%), sand (8.94 ha) and seagrass + sand (31.85 ha).

Keywords

• lyzenga algorithm, sentinel-2a, benthic habitat, salarangan island