Research on Coastline Automatic Extraction Methods Based on Remote Sensing Images

In view of the characters of mouth area, such as wide area shallow muddy water and mud deposit coast, morphological automatic algorithm based on wavelet transformation is introduced. The first step of this method is differentiating vegetation from soil effectively by spectrum spatial revolution, with linear transformation it is done in multi spectrum space formed by remote sensing images. Image noise is eliminated with multi scale and direction image noise elimination method which is based on wavelet transformation. Coast line is extracted rapidly, with combination abstracted mud spatial distribution data in shallow sea water, in this step we use image divide automatic seek seed point, rapid increasing algorithm, smooth-edged, edge tracing methods etc. The algorithm we used has the following advantages, such as computing rapidly, high automation, small personal effect, it supplies effective comparison coast line data for the continuously analysis.