There is a need to assess the quality of satellite data and images. When a
satellite produces data products for a particular application, their quality must
be verified before this data is delivered to the user community. On the user side,
after applying application algorithms to the data or images, the processed
images need to be assessed to make sure that the derived intermediate or final
products meet the requirements of the application. This chapter describes image
quality metrics used to assess the original satellite data products and the derived
The reason for defining a set of comprehensive image quality metrics is
obvious. An optical satellite sensor suffers from degradations in the acquisition
process related to instrument characteristics, for example, radiometric noise and
modulation transfer function (MTF). Different degradations introduced by the
acquisition system cause a loss of image quality. The first degradation on the
produced image products is radiometric noise caused mainly by photonic
effects in the photon detection process, by electronic devices, and by quantization.
This noise can often be assimilated to white noise even if some correlation exists
between different bands. A quality metric called the signal-to-noise ratio (SNR) is
often used to quantify how much the signal has been corrupted by noise.
Other degradations are due to the optical characteristics of the spectrographs.
The point spread function (PSF) can cause a smoothing effect along the spatial
dimension. The dispersion element of the spectrometer and the characteristics of
the detector array can produce a smoothing effect along the spectral dimension.
During the characterization of an optical satellite sensor, properlymeasuring the
image quality related to specific application needs using the quality metrics can
help enhance the performance of the sensor by focusing the crucial characteristics
to be improved.