Coastline Extraction on Remote Sensing Image Using Gaussian Process Classification

For more rapid and accurate coastline extraction methods, this paper presents a method of coastline extraction on remote sensing image using Gaussian process classification and then the coastline in coastal waters of Beihai city in Guangxi has been extracted. The results show that coastline extraction on remote sensing image using Gaussian process classification overcomes a series of public problems about coastline extraction that over-learning and the limitations of poor generalization ability in small sample exist in Artificial Neural Network and hyper-parameters are difficult to choose for Support Vector Machine. At last, the success of coastline extraction provides a efficient computing method for measurement and identification of coastline.