Lee, D.H.; Acharya, T.D., and Lee, H.S., 2017. Evaluation of seabed classification using hyperspectral data around the coastal area of Geoje Island, Korea. In: Lee, J.L.; Griffiths, T.; Lotan, A.; Suh, K.-S., and Lee, J. (eds.), The 2nd International Water Safety Symposium. Journal of Coastal Research, Special Issue No. 79, pp. 324–328. Coconut Creek (Florida), ISSN 0749-0208.
Seabed classification is an important part of current coastal research because it characterizes the seabed and its habitats. Seabed characterization makes the link between the classified area and the physical, geological, chemical or biological properties of seabed. This paper addresses the possibilities of the use of the hyperspectral data obtained from a CASI-1500 airborne sensor to map the seabed covers around the coastal area of Geoje Island, Korea. After radiometric, geometric and atmospheric correction for the raw hyperspectral data, the classification was performed in three steps. Firstly, thirteen classes of seabed were identified using an unsupervised spectral angle mapping algorithm in combination with data collected by ground survey. Secondly, seabed mapping was performed for each class separately. Finally, an accuracy assessment of the seabed mapping results was performed using data from ground survey. The overall accuracy was 84.29% with a kappa coefficient of 0.828. The results indicated that the hyperspectral data can help not only to classify the seabed material remotely and precisely, but also to construct the continuous geographical information for an effective management and conservation of the coastal area in Korea.