Quality Assessment of Pan Sharpened Images Using Object Based Image Segmentation

Earth observation satellite sensors might exhibit either high spectral or high spatial characteristics data. Image pan-sharpening or data fusion is widely used method for increasing the spatial resolution of multispectral images. Hence, the task is to enhance the spatial resolution of the multispectral images by merging them with spatially more highly resolved panchromatic data obtained from a satellite sensor. However, the quality evaluation of the final pansharpened image is a difficult task in the field of remote sensing, since a reference image is usually unavailable. This paper emphasizes the analysis of spectral and spatial quality of pan-sharpened images using object based image segmentation technique. The comparative study indicates that IHS+NSCT based image sharpening method provides better performance for quality evaluation of pan-sharpened images.