ibul 2004 
spikes in 
> concept 
image to 
«e 1998). 
:hromatic 
ing Scale 
gery. 
t distort a 
message. 
le remote 
from the 
ments on 
ormation 
tio, noise 
perature, 
the root 
'oduces a 
its. This 
ved. This 
measured 
> level of 
referred 
reported 
e radiant 
yduce an 
at which 
004). 
LAT): in 
iat yields 
)4). 
surement 
h a two- 
itude and 
e spatial 
ationship 
ag the 
)04). The 
Transfer 
measure 
source of 
ource of 
function 
int target 
nance of 
(CCRS, 
dulation 
ve edge 
l. 
ric (and 
orizontal 
jc filed 
urface of 
arization 
g object 
specially 
spectrum 
| Stokes 
   
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B1. Istanbul 2004 
  
parameters as important adjunct attributes for characterizing 
polarization in remote sensing systems. 
polarization sensitivity: the magnitude of change in 
performance of the remote sensing system as a result of 
polarization effects. 
precision error: see random error. 
presence of artifacts and flaws: artifacts and flaws within 
imagery can include items such as line drop outs, bad pixels, 
banding, distortion, etc. that appear within an image and affect 
its usefulness. 
pulse target: difference between maximum average value and 
background area brightness level divided by the 1-sigma SD 
noise estimate of the background areas. 
quantization: the process of converting continuous values of 
information to a finite number of discrete values. It is expressed 
as a number of bits. A 10-bit quantization means that the 
measured signal can be represented by a total of 210=1024 
digital values, say from 0 to 1023. (Kramer, 2001) The 
quantization level of an imaging system should be sufficient to 
meet the SNR requirements of the application. 
radiance: radiant flux per unit area per unit solid angle. 
radiant flux (radiant power): the total amount of power in a 
defined optical beam, measured in watts 
radiometrically accurate IFOV (RAIFOV): the resolution for 
which the MTF is higher than 0.95. 
random error: refers to the unpredictable, non-correlated 
component of the total error (Coleman and Steele 1999), 
calculated using mathematical laws of probability (Wolf, 1983). 
Also known as precision error. 
relative edge response (RER): (a) a measure of a system's 
ability to distinguish a straight, high contrast edge; (b) special 
case of the ESF, where the ESF is normalized. (c) a geometric 
mean of normalized edge response differences, between zero 
and 1 (the dark side of the edge at zero and the bright side at 
unity), measured in two directions of image pixels (X and Y) at 
points distanced from the edge by -0.5 and 0.5 GSD. 
relative geo-locational accuracy: (a) a measure of the integrity 
of angular and distance relationships of point features within an 
image; (b) the placement of point features relative to other 
points, sometimes called point-to-point accuracy. Relative 
accuracy is often computed as a function of distance. By 
accounting for relative error, an image having a large 
systematic error, or bias, may be shown to conserve positional 
relationships among features and thus, can be shown to have 
value for some applications (Ager 2002). 
relative radiometric accuracy: the difference between 
measured radiometric values from pixel-to-pixel and/or from 
band-to-band. It is the accuracy that is internal to the 
instrument, and not relative to an absolute standard. 
signal-to-noise ratio (SNR or S/N): (a) the amount of contrast 
between the bright and dark points of a 2-dimensional target 
divided by the variance in the homogeneous regions; (b) the 
ratio of the level of information-bearing signal power to the 
level of noise power (Kramer 2001). SNR can also be defined 
as the ratio of the power in a desired signal to the undesirable 
noise present in the absence of a signal (Photonics Directory 
2004). 
spatial out-of-field: The response of a system to radiation 
outside the field-of-view of the system, which is usually 
associated with stray light. This is often a measure of the wings 
ofthe PSF. 
spatial resolution: ability of a sensor to resolve spatial objects 
spectral bandwidth: For a given spectral band, the spectral 
bandwidth is the width of the spectral response curve at the full- 
width at half maximum (FWHM) value of the spectral response. 
spectral responsivity: a measure of a system's response to a 
known electromagnetic radiation field at a specified 
wavelength. The spectral response depends on the spectral 
characteristics of the detector, filters, telescope, and other 
components. Spectral response as a function of wavelength can 
be modeled as the spectral product of the response of the 
components, such as the detector, with the spectral transmission 
of all filter components, and the diffuser (National Physics 
Laboratory, 2004). See also spectral bandwidth, central 
wavelength, and cross-talk as additional key parameters 
related to responsivity. 
standard uncertainty: an estimated standard deviation. 
Stokes parameters: the parameters, relative to polarized light, 
that are usually represented as: I, the intensity of the light beam; 
M, that part of the beam polarized in the horizontal plane; C, 
that part polarized in the +45° direction; and S, that part 
circularly polarized (Photonics Directory 2004). 
subjective parameters: when evaluating image quality, there 
are often attributes that are not traceable to SI Units. See 
interpretability, general image quality function, and 
national image interpretability rating scale, presence of 
artifacts and flaws, and suitability to applications. 
suitability to applications: an image's suitability to various 
research and operational applications must ultimately be 
considered. Suitability refers not only to quantitative 
radiometric, spatial, and geometric performance of a sensor, but 
also to the overall ability of the acquired data (imagery) to 
address the needs of users for specified application. This can 
also be referred to as validation. 
systematic error: a measurement error that follows some 
mathematical or physical law; also known as correlated error, or 
bias. If the conditions causing the error are measured, a 
correction can be calculated and the systematic error 
eliminated. Systematic errors will remain constant in magnitude 
and algebraic sign if the conditions producing them remain the 
same (Wolf 1983). In this case the systematic error is the fixed 
or constant component of the total error (Coleman & Steele 
1999). 
systematic noise: noise created by non-random sources such as 
non-uniformity in the detector response caused by dead 
detectors or noisy pixels. This type of noise can produce image 
artifacts such as striping and banding in the image.