Coastline Identification from Remote-sensing Image Using Informative Vector Machine Learning

摘要

For a more rapid and accurate coastline identification method, this paper presents a method of coastline identification on remote-sensing image using Informative Vector Machine and then the coastline of Beihai Silver Beach in Guangxi has been identified. The results show that method of coastline identification on remote-sensing using Informative Vector Machine avoids the accuracy deficiency of the edge detection and threshold segmentation, and overcomes a series of open problems that difficulty in determining the optimal network topology structure in artificial neural network and the hyper-parameters in Support Vector Machine. The quick and accurate identification of coastline using the method has been realized which provides an efficient mean for the identification and monitoring of coastline.