| Popis: |
Both spatial and spectral information is used when a hyperspectral image is modeled as a tensor. However, this model does not consider both the class and within-class information about the spectral features of ground objects. This means that further improving classification is very difficult. The authors propose that class information, within-class information, and pixels are selected to model a third-order tensor. The most important advantage of the proposed method is that all the pixels of one class are mapped to the same coefficient vector. Therefore, the within-class scatter is minimized, and the classification is improved when compared to the previous methods. |
