ınbul 2004
ski, 2003;
has been
0 km over
M derived
one of the
js a new
'apabilities
so true for
iments) all
vels) are
listribution
ent which
nd market
RS images
for other
ns.
nal use of
ucts, some
d be given
The Earth
osed to set
especially
nstrument.
a case by
noting and
1g that the
1 be done
denver in
of ISPRS
ent ISPRS
lative was
published
0 potential
aper, is in
chaired by
for ISPRS.
Christophe
age (Marc
y and IGN
nce Data
nsible for
reas. This
e expected
M (derived
icy) and/or
better than
stereo pairs
or several
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B1. Istanbul 2004
methods) and to evaluate them in terms of quality and
accuracy, and when possible to compare (or combine) them
with other sources (ex HRG, SRTM, laser...).
8. THE PI SELECTION
After the publication of the HRS-SAP initiative 34 proposals
were received from all parts of the world. A Selection Team
was organized on March 19, 2003 in Toulouse, with the HRS
Secretariat and the ISPRS Secretary General.
Each proposal was quoted according to several criteria. Due
to the very short schedule only areas where HRS data was
already available were accepted. Then the expected quality of
the proposed Reference Data and also the experience and
scientific references of the candidates were also quoted.
From the 34 proposals 11 have been selected. The candidates
were informed of this selection. When pre selected they have
been asked to formally agree the HRS SAP rules and to send
the proposed Reference Data for checking.
Among those pre-selected PI's most of them have agreed and
signed the PI Agreement, and sent their Reference data to the
Secretariat. Unfortunately an interesting site on Gujarat-
India, proposed by Dr. Ramesh P. Singh (Department of
Civil Engineering, Indian Institute of Technology, India)
could not be confirmed, as the Reference DEM has been
found not accurate enough. Another site in the Alps was also
cancelled due to the fact that the pre-selected PI, (Etienne
Berthier, GDR STRAINSAR, France) could not accept the
proposed HRS images due to too large snow coverage.
Fig 6 HRS stereo pair on test site 5 (Aix-en-Provence area, France)
Then nine sites were confirmed : three in France: Manosque,
Aix-en-Provence (Fig. 6) and Montmirail; three in Europe,
outside trance: Chiemsee (Bavaria - Germany), Liege
(Belgium), Barcelona (Catalognia- Spain); and three in other
parts o the World: Merowe (Sudan -Africa), Melbourne
(Australia) and Rasht (Iran). Unfortunately, no site could be
selected in America The selected sites are well diversified in
terms of climate, relief or landscape and it is hoped that
future results could be representative of most situations in the
world.
9. CO-I SELECTION
Few days after the pre-selection of the PI's in March 2003 a
call for participation as Co-I was published by ISPRS. Then
19 new proposals have been received by the end of May
2003.
The selection was made using email exchanges between the
Selection Team, asking when necessary more information to
the candidates. For this selection the main criterion was the
scientific and technical professional capacity (in DEM
production) of the proposed team.
After this selection, made on June 14, 2003, 13 Co-
Investigators have been agreed and all of them signed the
requested Co-I Agreement. A 16" Co-Investigator has been
selected in November 2003, after having formalized his
proposal during the ISPRS Workshop in Hanover. As some
Col’s were already PI's and some have been selected on two
test sites the total number of experiments is 29.
10. DATA DELIVERY
In order to avoid unnecessary data duplication all HRS
images and Reference Data (Table 7) were produced and sent
at the same time, by mid July 2003, by Spot Image.
Test site HRS Reference Data
stereo
pairs
1 Montmirail 1 DEMs Im
2 Merowe 1 DEM 1m, GCPs
3 Liege 1 DEM 1m, 74 GCPs
4 Melbourne 1 DEM 1m, 33 GCPs
5 Aix-en-P. 1 DEMs 0.5-1m
7 Rasht 1 DEM 1.6m, GCPs
8 Barcelona 2 (+2 | DEMs |. Im
HRG) Orthoimages
9 Bavaria 2 DEMs 0.5-2m, 82
GCPs
| 10 Manosque 2 DEMSs 0.5-1 m
Table 7 Data provided to the investigators
It can be noted that on the Barcelona test site two HRS stereo
pairs have been provided, and also two HRG images, at 2.5m
resolution, giving the opportunity to test tri-stereo.
11. INVESTIGATORS
For each test site the number of investigators is between two
and seven:
- 2 for Merowe (Sudan), Liege (Belgium), Melbourne
(Australia), Aix-en-Provence;