VALS 
LY - 
rocessing, 
tion from 
ny remote 
d in-field 
‘entre, the 
servation 
> Imaging 
ms supply 
properties 
unction of 
geometry. 
inally, the 
to images 
impinging 
039, 9, ) 
direction 
ering the 
ffuser for 
ssed as a 
994; Qiu, 
| diffuser, 
it obvious 
? specular 
e spectral 
observed 
angles are 
iny rough 
eflectance 
ation and 
> azimuth 
assuming 
e for off- 
hik et al., 
study the 
e, that is 
stribution 
| both the 
heoretical 
lating the 
ibution to 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B1. Istanbul 2004 
  
Intense laboratory and in-field experimental activity have also 
been carried out in many research institutes for better 
characterizing BRDF properties of natural targets. Let us 
remind the European GOniometric Facility (EGO) of the Joint 
Research Centre (at Ispra, Italy) (Sandmeier et al., 1998), the 
LAbor-GOniometer System (LAGOS) and Fleld-GOniometer 
System (FIGOS) of the University of Zurich (Sandmeier et al., 
1999), the Portable Apparatus for Rapid Acquisition of 
Bidirectional Observation of Land and Atmosphere 
(PARABOLA) instrument of NASA — GSFC (Deering and 
Leone, 1986), and an instrument deveopled at Miami Univerity 
(Florida, USA), which is able to perform simultaneously 
multiple viewing-angle measurements (Voss et al., 2000). 
Recent satellite sensors such as the Multi-angle Imaging 
SpectroRadiometer (MISR) on the Earth Observation Science 
(EOS) Terra platform (Diner et al., 1998) and the Compact High 
Resolution Imaging Spectrometer (CHRIS) on board of 
European Space Agency (ESA) Proba platform (Cutter et al., 
2003) supply experimenters with their off-nadir tilting 
capability. 
Following this general trend aimed to improve the current 
understanding of directional properties of reflection from a 
surface, we show laboratory multiangular observations of 
natural sands obtained with a custom instrument whose main 
properties are presented in Section two. Section three describes 
the calibration procedure to take into account the BRDF 
properties of reflection from a surface as measured by our 
instrument. Some preliminary results are presented in Section 
four and open problems and conclusion are drawn in Section 
five. 
2. SYSTEM CONCEPT 
Although the BRDF is an important parameter for describing 
the surface reflectance, its measurement is hindered even for 
simple surfaces from the impossibility of yielding field-of-view 
(FOV) having a vanishing width. 
Moreover, because this function varies versus both illumination 
and viewing angle, many measurements are required. Therefore 
we have developed a suitable goniometric head for the ZEISS 
MCS 501 fiber optics spectrometer, whose characteristics are 
listed in Table I. 
  
Two 600 pm (core diameter) 
Optical entry: fibers with NA~0.25 for 
illuminating and viewing 
  
Holographic concave grating 
g element: 
= 
© 
Dispersin : 
with 157 grooves/mm 
  
Hamamatsu with 1024 
Detector array: 
elements 
  
215 nm — 1015 nm (nominal) 
Spectral range: 
280 nm — 900 nm (working) 
  
Spectral resolution: 2-3 nm 
  
Internal 75W Xenon lamp 
Illuminating source: 
(CLX 500) 
  
Digitalization: 16 bit 
  
  
  
  
Table I. Spectral and radiometric properties 
of the ZEISS MCS 501 spectrometer utilised 
for laboratory reflectance measurements 
Each fiber, which is employed for illuminating the surface and 
collecting the reflected radiance, is terminated with GRIN-rod 
lenses SLW30 to collimate the outgoing and incoming beams. 
In the actual configuration the illuminating and reflected beams 
have a divergence of 108 mrad and a spot size diameter (at the 
end of the GRIN-rod) equals to 1.5 mm. Figure 1 shows a 
picture of the optical head employed to cary out BRDF 
measurements. 
  
Figure 1. Picture of the optical head employed for laboratory 
BRDF measurements 
The optical fibers are arranged on a goniometric mounting so 
that the centre of the illuminated spot remains fixed on the 
sample surface while moving the source fiber at a different 
illumination angle. The only change of the illuminated surface 
is connected with the deformation of the source spot, which 
becomes elliptical with increasing the viewing/illumination 
angle. 
We have performed measurements for illumination (zenith) 
angles of 0°, 15°, 30° and 45°, mapping the bi-conical 
reflectance for different viewing angles in the source principal 
plane. Due to mechanical constraints position of illuminating 
and viewing fibers have been limited are limited to zenith 
angles not greater than 60°. 
3. CALIBRATION PROCEDURE 
The geometrical properties of a reflecting surface are readily 
described by its BRDF, denoted symbolically as 
P gnpr 9 99.09,9,,0,), which is defined (Nicodemus et al., 
1977) as the ratio of the radiance dL" (4,95,09,9,,0,; E;) 
scattered into the direction (9,,0,) to the irradiance 
dE,(,9,,09) impinging at angle (85,05) ona unitary surface 
area (see Figure 2 for a coordinate description): 
di" (4 9 909,.0,: ,) 
P sapr O599,09.9,.0,) 7 (2) 
dE; (4.,99,0,) 
  
Due to its definition BRDF is a density of reflectance [sr] and 
it can take values from zero to infinity. Let us note that the 
BRDF, defined as ratio of infinitesimals (vanishing quantities), 
is a derivative with instantaneous values that can not be directly