Using the Mean-shift and Region Growing Algorithm to Extract Coastline : A Case Study of the Arctic Region of Wrangel Island

In recent years, the activity of the human has been more and more continual which caused the change of the coastal zone more active than before. So it is vital significant to accurately and quickly extract coastline and real-time monitor the changes. Many approaches have been developed to automatically extract coastline. After comprehensive comparison, it was found that combining mean-shift with region growing algorithm could extract the coastline very well. In this paper, the MNDWI was firstly used to roughly determine the range of shoreline. Then, mean-shift was applied to remove the noise, and finally the continuous coastline was extracted using the method of regional growth. Experiments show that the method could extract the continuous coastline accurately and quickly.