| Local Binary Graph Feature Reduction for three-dimensional Gabor Filter Based Hyperspectral Image Classification |
| Paper ID : 1152-SMPR |
| Authors: |
|
Mohsen Darvishnezhad *1, hassan ghassemian2, maryam imani3 1Electrical Engendering 2Tarbiat modares University 3Tarbiat modares university |
| Abstract: |
| The classification of a hyperspectral image order amounts to knowing which pixels include different spectrally different elements that have been defined by the user. One of the challenges of the hyperspectral image classification is the fusing spectral and spatial features. There are several methods for fusing features in hyperspectral image classification. Three-Dimensional Gabor Filters are the best methods for simultaneously using spectral and spatial features. However, one of the problems with the use of the 3D Gabor filter is the high number of extracted features. The goal in this paper is reducing extracted features from 3D-Gabor filters to increase the classification accuracy in hyperspectral images. Therefore, in this paper in order to reduce extracted features, the Local Binary Graph (LBG) method is used. So, the proposed method results in reducing extracted features from the 3D-Gabor filters by LBG method. Finally, the experiments show that the proposed method improves increased to 96.2, 92.6 for Pavia University and Indian Pines dataset |
| Keywords: |
| classification, hyperspectral, spectral, spatial, feature fusion, Three-Dimensional Gabor filters. |
| Status : Conditional Accept (Oral Presentation) |
