Hyperspectral Target Detection Improvement Based on Noise Reduction and Spectral-Spatial Methods Composition

In order to providing the accuracy of spectral target detection in this research, we use three different procedures. They are decreasing noise and spectral correlation of Hyperspectral images by the use of algorithm MNF(Minimum Noise Fraction), compounding the detection algorithms by the use of ANFIS(Adaptive Neuro-Fuzzy Inference System) method and finally the spectral-spatial target detection in which we use the spatial correlation of data along the spectral data. We assessed the above mentioned procedures on two cases: initial image and decreased noise-spectral correlation image. The result of quantitative and qualitative assessments of tests showed that through the improvement detection procedures in this research, the procedure of compounding detection algorithm on the initial image and the assessing algorithms of decreased noise-spectral correlation image have respectively more and less effect on providing the target detection accuracy.