Research on remote sensed image Classification Based on morphological feature extraction

With the development of remote sensing technology, remote sensed data provide detailed information of structure and spectral about ground scenes. However, the traditional classification methods only take objects’ spectral information into consideration but neglect their structural information. There are two major reasons for this: First, the multi-spectral data contain a lot of information on the spectral properties of the land cover in the data, but no spatial information is inherent in the spectral data. Second, the individual images from a ground scene contain spatial information but very limited information on the spectral property of the data. So it is necessary to include both spectral and structure information in the process of multi-spectral data classification.