You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

New perspectives to save cultural heritage
Altan, M. Orhan

P. Ardissone (a) , P. Boccardo (a) , E. Borgogno Mondino (a) ,
(a) Politecnico di Torino - Dipartimento di Georisorse e Territorio
Corso Duca degli Abruzzi, 24- 10129 Torino - Italy
Email: paolo.ardissone@polito.it, piero.boccardo@polito.it, enrico.borgogno@polito.it
Commission V, WG V/5
KEYWORDS: cultural heritage, remote sensing, digital image processing, GIS, hyperspectral, warping, orthoprojection
This paper concerns the digital image processing of hyperspectral airborne data, devoted to landscape and cultural heritage analysis.
The project completes a three years GIS case study devoted to the characterisation of a no more visible middle age landscape. The
study area is located in the southern part of the Piedmont region (North-West Italy) and in particular in a valley whose height ranges
from 600 m to 2500 m, called Valle Po.
The aim of the past analysis was to integrate different thematic data such as archaeology, geology, botany and cultural anthropology
using a medium scale cartography (up to 1:5.000) and high resolution satellite images to georeference the above mentioned data.
During the last year, a hyper-spectral survey, conducted with the M1VIS airborne scanner, has been orthoprojected and used to derive
different thematic maps which give information related to:
• archaeological evidence;
• mine and quarry positions;
• lithotype characterization;
• vegetation cover and types.
All these thematic maps can be derived exploiting the 102 different spectral bands of the MIVIS sensor. Particular attention has been
paid to the calibration of the sensor, extraction of the spectral signatures and to the georeferencing of the above mentioned images.
Such operation was done orthoprojecting the images through a general sensor model procedure, self-implemented within the IDL
programming language. At the moment the group is also working to the development of a rigorous sensor model (attitude and
positioning parameters estimation) with the aim of improving the accuracy. Obtained thematic maps have then been linked to a
relational database aimed to query and correlate different and previously disjointed information.
The proposed procedure completes and integrates the project
“Landascape heritage and resource management: an integrated
information system of the Marchesato di Saluzzo” which was
aimed at studying the settlement development and the use of
land in the Po valley between X and XIV Centuries. During the
last three years our work was based on archive and bibliography
research with the aim of finding all the available information on
the “material structures”, such as castles, churches, settlements
and roads from the Middle Ages. Subsequently, a field survey
was carried out to locate these structures and to identify which
have disappeared and which have remained. All the data was
entered and structured in a GIS database together with most of
the other available data about the Po valley area: technical,
thematic, geological and botanical.
Over the last year the data queries and spatial analysis led us to
select test areas on which to conduct an MIVIS hyperspectral
airborne scanner acquisition. MIVIS data analysis has permitted
a close examination of the selected areas to look for buried
structures, exploit indications offered by radiometric and
spectral anomalies on the ground.
It is worth to here underline that the goal of the current project
is not only to increase knowledge about the past, nor simply
offer a significant aid to historical and archaeological studies
and data management, but also to provide aid to local
administrations for correct environmental management and
accurate cultural heritage safeguarding of the area.
Investigations have been carried on using:
a) a numerical Technical Regional Map (scale 1:10 000);
b) a 50x50m grid Digital Elevation Model;
c) Archaeological, geological and botanical available
data (derived from the “Marchesato di Saluzzo " GIS);
d) a MIVIS (Multispectral Infrared and Visible Imaging
Spectrometer) image of the Po valley mouth, acquired
on 18th December 2002, with a ground resolution of
about 4 meters.
The MIVIS hyper-spectral scanner is a modular instrument
composed of four spectrometers which simultaneously measure
the electromagnetic radiation from the Earth’s surface by
recording 102 spectral bands:
20 in the visible spectral region (0.43-0.83pm)
8 in the near infrared one (1.15-1.55pm)
64 in the middle infrared one (2.0-2.5pm)
10 in thermal infrared (8.2-12.7pm).
When dealing with territorial applications it is always important
to correctly approach the scale mapping problem. This means
that ground object positioning must be coherent for all the used
data (often coming from different sources and reference map
systems). Such problem can be easily solved for geocoded data
such as the ancillary and the cartographic ones. Not so easy is to
face the problem of the geocoding of MIVIS data maintaining
the ground position accuracy within an acceptable tolerance