01. Research for creating a land cover current status map with the medium scale of sea shoals using WorldView-2 imagery

Abstract

Sea shoals play a crucial role in socioeconomic development, security, and defense protection. In this paper, the authors use WorldView-2 satellite images to create a medium-scale current status map of the shoals in the sea. The classification method for the current state of land cover on WorldView-2 images is a combination of object-oriented classification and SVM (Support Vector Machine), a supervised classification. Thuyen Chai beach on Truong Sa island is the experimental area. Satellite image data is preprocessed and classified according to a standardized procedure. The overall classification accuracy is 92.45 %, and the Kappa coefficient is 0.90. The classification of the current state of the land cover on the WorldView-2 satellite images is used as the input database that creates the 1:50,000-scale current status map.