A note on spatial resolution measurement and its implications for image radiometry

Abstract

The smallest sized detail on the ground actually recorded by an airborne or satellite camera system can be different to that expected by an applications scientist or image analyst. When this misconception occurs, it may be due to the method used in determining the spatial resolution of the camera system concerned. The role of an optical system is of primary importance in setting a limit on the spatial resolution achieved by either a photographic or optoelectronic/digital camera system. This is because it is the first component of the imaging system to intercept radiation coming from an object on the ground. In circumstances where the smallest image patch produced at the focal plane by an optical system is comparable or larger in size than the pixel size/ spacing of a CCD then spatial resolution will become limited by the optics, not the CCD. When this occurs, the radiometric response of the camera system to small features on the ground will correspond to the small optical image patch, and not the size of pixels in the CCD. The spatial resolution characteristics of a camera system will usually be given in terms of the ground sampled distance (GSD), or as a ground resolved distance (GRD). The results obtained by the two methods of measurement these values represent are different, and the difference can be significant because of the sampling principles that are the basis of each method.

Digital camera performance where spatial resolution is determined by the optical component

“…T he image quality produced by a digital camera is frequently described in terms of the spatial resolution quoted by the camera manufacturer or by the vendors of imagery; this feature is usually related to a value at or close to the number or size of pixels used to record the image produced initially by the optical system. Attention has already been drawn to the uncertainties inherent where the pixel alone is used as a method of assessing image quality (Fisher, 1997; Cracknell, 1998; Thomson, 2009). Furthermore, the pixel‐based method introduces a fundamental flaw in any measurement, since it ignores the effect on image quality resulting from the optical system.…”

The Practical Effect of a Diffraction‐Limited Image for Photogrammetry

“…A ttention has previously been drawn to the cardinal role played by the optical system in the formation of an image (Thomson, 2009; 2010). The question is then raised of the practical effect that an optical limitation can have on image quality, when the smallest image patch possible is that determined by the optical system.…”