COMPARISON OF DIFFERENT FUSION ALGORITHMS IN URBAN AND AGRICULTURAL AREAS USING SAR (PALSAR AND RADARSAT) AND OPTICAL (SPOT) IMAGES

Resumo

Image fusion techniques of remote sensing data are formal frameworks for merging and using images originating from different sources. This research investigates the quality assessment of Synthetic Aperture Radar (SAR) data fusion with optical imagery. Two different SAR data from different sensors namely RADARSAT-1 and PALSAR were fused with SPOT-2 data. Both SAR data have the same resolutionsand polarisations; however images were gathered in different frequencies as C band and L band respectively. This paper contributes to the comparative evaluation of fused data for understanding the performance of implemented image fusionalgorithms such as Ehlers, IHS (Intensity-Hue-Saturation), HPF (High Pass Frequency), two dimensional DWT (Discrete Wavelet Transformation), and PCA (Principal Component Analysis) techniques. Quality assessments of fused imageswere performed both qualitatively and quantitatively. For the statistical analysis; bias, correlation coefficient (CC), difference in variance (DIV), standard deviation difference (SDD), universal image quality index (UIQI) methods were applied on the fused images. The evaluations were performed by categorizing the test area into two as “urban” and “agricultural”. It has been observed that some of the methodshave enhanced either the spatial quality or preserved spectral quality of the original SPOT XS image to various degrees while some approaches have introduced distortions. In general we noted that Ehlers’ spectral quality is far better than those of the other methods. HPF performs almost best in agricultural areas for both SAR images.