Spatial resolution assessment from real image data

Abstract

The radiometric and spatial characteristics of a remote sensing system specify the image quality. The determination of the image quality is carried out on specific resolution targets. The point spread function (PSF) is an essential parameter for characterizing the spatial imaging properties of the whole system. We present an approach for PSF determination using high-contrast edges found in the typical urban scenes. It is assumed that the PSF is described by a Gaussian. Instead of the PSF the edge spread function (ESF) was determined. For the investigations RapidEye multispectral orthophoto data (L3a) has been used. Clear transitions between bright and dark patches have been selected in the imagery. The analysis is carried out in a horizontal and vertical direction only. Because of the spectral dependence of the reflection properties, these transitions were examined for all of the five bands of the same image data. This presentation describes a special tool to determine the data quality based on PSF estimations of RapidEye image products. It is also shown that the determination of the influence of different image processing algorithms is possible.