Remote Sensing Water Quality In The Yangtze River Delta Classification Technology

The application of the technology of satellite remote sensing on the real-time watching & forecasting of inshore seawater's quality is very important for national economy. In this article, we use five kinds of data for seawater's quality from the inshore delta region of the Changjiang River supplied by the MODIS satellite to do the research of clustering.This article is mainly about the following content: (1) developing the platform for the analysis and processing of the satellite remote sensing data, and making a introduction about its functions.(2) realizing the K-means clustering algorithm, and especially bringing out a improving K-means clustering algorithm based on the optimization of the beginning center in clustering.(3) realizing the fuzzy C-means clustering on the remote sensing data, which is proved much better than the K-means algorithm in both astringency and accuracy by experiments, and integrating both of them to a improving algorithm for avoiding the partial optimization.(4)bringing out a Multi-Resolution Fuzzy Segmentation Method based on the Pyramid graphic delaminating structure which is about the comparability both in the classes and between them to classify the remote sensing data clustering, and the method is proved by experiments better in the sharply changing parts than the Fuzzy C-means algorithm.