| Title Of The Magazine |
Proceedings of Dragon 2 Programme Mid-Term Results 2008 - 2010 |
| Year Information |
2010 |
| KeyWord |
|
| Authors |
JIA Kun, LI Qiangzi, TIAN Yichen, WU Bingfang, ZHANG Feifei, MENG Jihua |
| Intro |
In this paper, the fusion data of HJ satellite multispectral and Envisat ASAR VV polarization backscatter data is investigated for crop classification, and explores the applicability of VV polarization data for crop classification. HJ multi-spectral data can effectively classify crop types in the study area with the classification accuracy of 89.2% using Support Vector Machine classifier. But the field borders can not be recognized with HJ data and leads to misclassification. It can be significantly improved with fusion data of HJ multi-spectral and ASAR data with the 5 percent increase of the classification accuracy. The results indicate that fusion data can take full advantage of spectral information of HJ multi-spectral data and the structure sensitivity feature of ASAR VV polarization data, enhances the spectral difference among different classes and improves crop classification accuracy. ASAR VV polarization data is sensitive to non-agrarian area of planted field, and can effectively distinguish the field border. |
| Links To Magazines |
http://dragon2.esa.int/symposium2010/ |
English
中文