Dimension reduction with heuristic methods in hyperspectral images

Heuristic Methods are based on searching the solution space quickly to get optimal or approximately optimal solution. These methods are relied on the systematically neighborhood change in search area and generally used to achieve the optimal solution in a short time in high dimensional search space. Examining the data including large scale of information such as hyperspectral images and eliminating redundant features (bands) is quite important for computation time and target classification/detection performance. In this study, band selection as a dimension reduction procedure is employed to hyperspectral images using several heuristic methods (Differential Evolution Algorithm, Genetic Algorithm, Simulated Annealing and Variable Neighborhood Search) and then useful bands are selected. In the next stage, as a preprocessing step of band selection, different grouping methods (Mutual Information, Correlation Coefficient and Spectral Clustering) are utilized for forming groups based on similarity between bands. Final stage, superlative bands which are found after band selection are tested with a classification algorithm (Support Vector Machine). To test the proposed methods real hyperspectral images (KSC and Çatalca0202) are used. Test results indicate that presented methods can be applied in hyperspectral imagery successfully.