Abstract:

Multispectral remote sensing data is the data source for dynamic monitoring of sandy land. The based data is sandy land information from remote sensing data. Bands selection directly impacts information extraction and land classification. The author applied the bands correlation and the best index OIF index to select optimal bands combining spectrum characteristics of the research objects, finally band 3, 4, 7 were the best combination. This will be advantageous to see and interpret. The study on combination of bands provided strong techniques to support for the post-processing.