A Proposal for Automatic Coastline Extraction from Landsat 8 OLI Images Combining Modified Optimum Index Factor (MOIF) and K-Means

The coastal environment is a natural and economic resource of extraordinary value, but it constantly modifying susceptible to climate change, human activities hazards. Remote sensing techniques have proved be excellent for area monitoring, the main issue detect borderline between water bodies (ocean, sea, lake or river) land. This research aims define rapid accurate methodological approach, based on k-means algorithm, classify remotely sensed images in an unsupervised way distinguish body pixels coastline. Landsat 8 Operational Land Imager (OLI) multispectral satellite were considered. proposal requires applying algorithm only most appropriate bands, rather than using entire dataset. In fact, by suitable bands differences no-water (vegetation bare soil), more results obtained. For this scope, new index optimum factor (OIF) was applied identify three best-performing purpose. direct comparison automatically extracted coastline manually digitized one used evaluate product accuracy. very satisfactory combination involving B2 (blue), B5 (near infrared), B6 (short-wave infrared-1) provided best performance.