A simulation approach to the spectral quality of multispectral images enhancement

摘要

The accuracy of environmental research with the use of remote sensing methods depends upon the quality of image data. Hence the acquisition of images of high spatial and spectral quality simultaneously is of crucial importance. To resolve the problem of structural limitations of sensors that influence the quality of acquired images, it is necessary to integrate data from different sensors. Using both original and simulated high spatial resolution images, the authors proposed an effective methodology for integrating image data from aerial and satellite sensors, allowing the generation of sharpened images while maintaining the highest degree possible of the spectral features of photographed objects. At the same time, the article concerns the impact assessment of spectral resolution of high spatial images upon the spectral quality of an enhanced image in multisensor data fusion process with the use of the pan-sharpening method. It is of extreme importance from the viewpoint of precision farming, where it is necessary to have as many parameters as possible that characterise agricultural and forest land. Moreover, the aim is for such parameters to reflect the reliable spectral characteristics of the studied objects.