In: Wagner W„ Székely, B. (eds.): ISPRS TC VII Symposium - 100 Years ISPRS, Vienna, Austria, July 5-7, 2010, IAPRS, Vol. XXXVIII, Part 7B 
In: Wagnei 
VALIDATION OF THE REFLECTANCE CALIBRATION OF THE ADS40 AIRBORNE 
SENSOR USING GROUND REFLECTANCE MEASUREMENTS 
The relat 
[W/m 2 /sr/^ 
U. Beisl a ’ *, M. Adiguezel a 
d Leica Geosystems AG, Heerbrugg, Switzerland (Ulrich.Beisl, Muzaffer.Adiguezel)@leica-geosystems.com 
Commission VII, WG VII/1 
KEY WORDS: Multispectral, Camera, Radiometric, Calibration, Correction, Atmosphere 
Since at-se 
path radia 
from a blu 
flight direi 
options fo 
Subtractio 
proximate 
because th 
ABSTRACT: 
In the past years digital airborne imaging data have become an appropriate tool to perform remote sensing tasks. With Leica XPro, a 
processing software for the ADS 40/80 sensor is now available that allows calculating ground reflectance data during rectification. 
Furthermore, a BRDF correction of the individual flight lines is possible. Since the two processing options are without additional 
user input, they are also used regularly for mapping purposes to speed up the mosaicking process. In this paper, a validation of the 
atmospheric correction and reflectance calibration is shown, using data from an image block in southern Germany, which was used 
in the project "Evaluation of Digital Camera Systems" by the German Society for Photogrammetry, Remote Sensing and 
Geoinformation (DGPF). Compressed and uncompressed image data was available in 8 and 20 cm ground pixel size together with 
in-situ ground reflectance measurements of different tarps and natural targets at the time of overflight. Since in the compressed data 
set the targets were observed several times at different view and illumination angles, also results from the BRDF correction could be 
compared. The results show a reasonably good agreement between reflectances measured on the ground and those calculated from 
image data in the atmospheric correction process. Furthermore it could be shown that the BRDF correction option considerably 
reduces the reflectance difference between different flight lines within selected targets of the test area. The discussion shows the 
limitations of the method. Finally, improvements are proposed to further increase the accuracy. 
1. INTRODUCTION 
During the last years there has been a growing interest in the 
radiometric properties of high-resolution aerial images. An 
assessment of the radiometric properties of large format digital 
cameras has been done by Markelin et al. (2008). However, up 
to now, no operational processes for the handling of absolute 
radiometric image data are yet implemented in European 
National Mapping Agencies (Honkavaara et al., 2009). 
Attempts to perform an absolute radiometric calibration of 
frame images have been started for the DMC (Ryan and 
Pagnutti, 2009, Gonzalez-Piqueras et al., 2010) and the 
UltraCam sensors (Alvarez et al., 2010). Molina et al. (2010) 
are using an empirical polynomial approach to balance the 
different image brightness in a DMC image mosaic. 
The ADS40/80 system features an absolute radiometric calibra 
tion since 2006 (Beisl, 2006) and provides a workflow in the 
Leica XPro software for calculating ground reflectances (Beisl 
et al., 2008). A validation of the reflectance product is also 
shown in this symposium (Markelin et al., 2010). 
ADS40 data has been used in various remote sensing applica 
tions like avalanche prediction (Biihler et al., 2009), classifica 
tion of benthic habitats (Green and Lopez, 2007), canopy cover 
and tree species classification (Waser et al., 2008, Waser et al., 
2010), urban classification (Emmolo et al., 2008), and archaeo 
logical prospection (Kellenberger and Nagy, 2008). 
In this article, a validation of the radiometry of the ADS40/80 
reflectance and BRDF correction products is done using an 
image block from a rural area in southern Germany in order to 
assess the usability of the ADS40/80 system in remote sensing 
projects. This image block was also used in the project "Evalua 
tion of Digital Camera Systems" by the German Society for 
Photogrammetry, Remote Sensing and Geoinformation (DGPF). 
(Cramer, 2010a). 
2. MATERIALS AND METHODS 
2.1 Leica ADS40/80 sensor 
The Leica ADS40/80 sensor uses a line scanner principle, 
which is particularly suited for absolute radiometric calibration, 
since the focal plane is temperature stabilized, the lens aperture 
is fixed, and the CCD lines have a considerably higher dynamic 
range than any area-CCD. This way a single integration time 
setting can be used for a whole flight line with heavily varying 
surface reflectance without the need of changing the aperture. 
2.2 Radiometric processing with Leica XPro 
Since the dark signal non-uniformity and the photo response 
non-uniformity (lens falloff) are already corrected on the sensor 
head, the relation between raw digital numbers (DN) and cali 
brated digital numbers (CDN) is just a linear function with a 
single calibration factor Cj (camera constant) for each of the 12 
CCD lines in a sensor and the actually used integration time (IT 
[s]). This CDN product is the default output of the XPro recti 
fier (“no correction”). 
CDN = DN * 50* cjIT (l) 
In order tc 
in 2008 t! 
workflow, 
ance, and 
(RCDN) (1 
tance R is 
An additic 
the cross t 
ground sui 
strips calit 
2.3 Airb< 
In 2008 th 
tal Camen 
metric anc 
performan 
manual ste 
text of the 
Vaihingen 
and medii 
were flow 
2010a). TI 
gust 6, 2( 
(compressi 
DGPF dire 
Line 
GSD 
[cm] 
L6 
8 
L5 
8 
L4 
8 
L3 
8 
L1 
8 
L7 
8 
L2 
8 
L8 
8 
H5 
20 
H6 
20 
H4 
20 
H3 
20 
H2 
20 
H1 
20 
Table 1. ' 
(IT) for 
Bw=backv 
raw=uncor 
Corresponding author.