Klasifikasi habitat dasar berbasis objek di perairan dangkal karang lebar dan pulau lancang

The object-based classification technique (OBIA) is one of the benthic habitat mapping techniques besides the conventional (pixel-based) method. The mapping of the OBIA method using machine learning algorithms is limited to the waters of Karang Lebar and Lancang Island. This study aims to determine the performance of machine learning algorithms (support vector machine (SVM), decision tree (DT), random forest (RF), and k-nearest neighbor (KNN)) in classifying shallow water benthic habitats based on objects using Sentinel satellite data. -2. The classification method used is the OBIA method with two levels of analysis. A total of 6 benthic habitat classes were obtained from field observations and Agglomerative Hierarchial Clustering analysis, namely coral, rubble, seagrass, rubble sand, and sand. The results obtained include the first level separating land, shallow sea and deeper sea. The second level is classification using a machine learning algorithm, the results of the classification show that the SVM algorithm gets a higher accuracy value than other algorithms with an accuracy of 84% in Karang Lebar waters, then in Lancang Island waters it gets an accuracy of 80% with the SVM algorithm. The bottom habitat of the shallow waters of Karang Lebar and Lancang Island can be well mapped using the OBIA method. The difference in the level of accuracy between the waters of Karang Lebar and Pulau Lancang is caused by the level of turbidity of the waters.