Land cover classification and winter wheat mapping using data mining strategy

With the increasing sensors available, the knowledge gap between satellite data and application needs is continuously increasing. Aimed at the information demands for mapping land resources, a study on the key technologies of intelligent information processing for satellite remote sensing is presented based on data mining and knowledge discovery strategy as well as information theory, which is suitable for Land cover classification and feature extraction. Significant results have been achieved by performing a test on land cover classification in Minshan area, which is the most important ecosystem region of panda distribution in China, together with an experiment on extracting the winter wheat distribution and the status of agricultural structure adjustment of Beijing from multi-temporal Landsat images.