Benthic Habitat Mapping on Different Coral Reef Types Using Random Forest and Support Vector Machine Algorithm

摘要

Machine learning classification in remote sensing imagery is considered capable of producing classification results withhigh accuracy in short processing times. This research was conducted with the aim of mapping the spatial distribution ofbenthic habitat on different types of coral reefs using PlanetScope image with Random Forest (RF) and Support VectorMachine (SVM) algorithm in the waters of Flores Island, NTT. Benthic habitat information from field surveys were usedto train the RF and SVM algorithm and validate the classification results. The classification results indicated that MesaIsland, the Northern and the Western side of Labuan Bajo are dominated by seagrass beds, and on Bangkau Island isdominated by coral reefs and bare substratum. The highest overall accuracy of the RF classification results is 71.88% fromWest Labuan Bajo (fringing reef) result. Meanwhile, the highest overall accuracy of the SVM classification is 76.74%from Bangkau Island (patch reef) result.