Remote sensing technology provides a reliable means of creating an inventory of coastal natural resources to assit with their management. The purpose of this study was to discriminate seagrass density using medium and high resolution satellite imagery. The field survey was conducted on Bahuluang island, Spermonde archipelago.Imageryof Sentinel-2a as medium resolution (10 m) and SPOT-6 as hight resolution (1.5 m) were used to discriminate the seagrass. Both pixel-based (k-Means). Object-based classification was used as an approach of Nearest Neighbor (KKN) methods using segmentation method before doing classification, and in this methode used parameters that must be determined in example based feature extraction workflow using Edge scale level dan full lamda scale level algorithm, namely parameters of spectral, texture, and spatial. Pixel-based classification methods provideds better over all accuracy than object-based classification and can be used for seagrass mapping in the Spermonde Archipelago. However, it is necessary to conduct a more detail assessments at different water depths.