Increasing the accuracy of data extraction from OLI data using the FFT-IHS method

The aim of this study is to use of the FFT_IHS method to increase the accuracy of data extraction from OLI of Landsat 8 data. For this purpose, a window of OLI images of Ardabil County was selected and, after applying the necessary preprocesses include atmospheric correction, the multispectral and panchromatic bands were fused with the FFT_IHS method. In order to evaluate the capability of FFT_IHS method to increase the accuracy of information extraction, the training data were taken from the before and after applying this method. Correlation of training data’s was evaluated using the Jeffries Matusita index and the training data’s were classified into 8-classes using Support Vector Machine algorithm. The results showed that image classification before the fusion of bands has a overall accuracy of 88.3% and a kappa coefficient of 0.87 and after fusion with FFT_IHS, the overall accuracy is increased to 96.3% and the Kappa coefficient is to 0.96.