Beyond Pansharpening: Advances in Data Fusion for Very High Resolution Remote Sensing Data

Due to the high spatial resolution requirements especially for urban information systems, aerial photography has been used as standard imaging input. The advent of new satellites with a resolution of better than 1m (e.g., IKONOS, Quickbird) and digital airborne scanners with excellent geometric fidelity and high spatial resolutions in the cm range (e.g., HRSC, ADS, DMC, Ultracam), however, challenges the analog airphoto techniques. These airborne and spaceborne high resolution sensors offer an advanced potential for generating and updating GIS databases especially for urban areas. Highest resolution for all sensors is obtained in their panchromatic mode whereas the multispectral information is usually acquired at lower resolutions. The ratio between the panchromatic and the multispectral resolution is usually in the order of 1:2 to 1:8. Especially for vegetation analyses and urban sprawl quantification, multispectral information is a mandatory requirement. To obtain multispectral image data with high spatial resolution, multispectral and panchromatic images have to be merged (pansharpening). Image transforms such as the Intensity-HueSaturation (IHS) or Principal Component (PC) transforms are widely used to fuse panchromatic images of high spatial resolution with multispectral images of lower resolution. These techniques create multispectral images of higher spatial resolution but usually at the cost that these transforms do not preserve the original color or spectral characteristics of the input image data. The FZG has developed a new method for image fusion that is based on the standard IHS transform combined with filtering in the Fourier domain. This method preserves the spectral characteristics of the lower resolution images. Using this spectral characteristics preserving fusion method, multispectral images can be pansharpened without changing their spectral characteristics.