ISPRS Commission III, Vol.34, Part 3A ,,Photogrammetric Computer Vision“, Graz, 2002
OSIRIS: A SIMULATOR OF OUTDOOR SCENES IN THERMAL INFRARED RANGE
T. Poglio ®° *, E, Savaria?, L. Wald?
* Alcatel Space Industries, System Architecture Division, 06156 Cannes, France —
(thierry.poglio, eric.savaria)@space.alcatel.fr
e Groupe Télédétection & Modélisation, Ecole des Mines de Paris, 06904 Sophia Antipolis, France —
(thierry.poglio, lucien.wald)@ensmp.fr
KEY WORDS: Imagery, Landscape, Modelling, Remote Sensing, Temperature, Three-Dimensional
ABSTRACT:
This paper deals with the simulation of very high spatial resolution images in the thermal infrared range, from 3 to 14 um. It recalls
the physical processes occurring in this spectral range. The complexity of the simulation of such processes is explained. The
specifications of the simulator and a methodology for the simulation of infrared imagery are proposed. This methodology enables a
very accurate simulation of the signal coming from each object constituting the landscape. The interactions between the radiations
and objects and the convective and conductive interactions between objects themselves are considered. Their variations in time, and
the recent past of the temperature and the humidity for each object, are taken into account. To reproduce these physical phenomena,
the computation is performed on elements, which are defined as homogeneous entities with respect to the physical processes. The
architecture of the simulator OSIRIS using the elements is presented. Finally, examples are shown, showing the efficiency of the
simulator and the methodology used.
1. INTRODUCTION
There is a growing concern of very high spatial resolution
imagery in the infrared range from 3 to 14 um. Simulators of
landscapes in thermal infrared range need to be developed. One
of the major points in the simulation is the accurate knowledge
of the input parameters, and particularly the accurate knowledge
of the scene to be observed. Synthesis approaches to outdoor
scenes offer the possibilities to simulate changing
meteorological conditions, different places, different
landscapes, different times and different spectral bands.
As physical processes play an essential part in the signal coming
from the scene, there is a need for research on spatial, spectral
and temporal properties of natural and human-formed objects,
their image signatures and their modelling. The synthesis of
outdoor scenes in the infrared range is at the crossings of
computer vision, physics and remote sensing.
In thermal infrared, the flux coming from an object is partly
emitted by the object because of its own temperature, and partly
due to the reflection of incident rays on the surface of this
object. Depending on the surface material and the spectral band,
emission or reflection process dominates the signal. For each
object in the scene, the landscape simulator predicts the heat
exchanges between objects, the temporal evolution of heat
balance, the surface temperature, the spectral emission and the
spectral reflection of all incident fluxes. Such a simulator takes
into account 3-D landscape description, environmental
conditions, thermal and optical characteristics of the objects,
and the spectral band of the sensor.
Jaloustre-Audouin (1998) and Jaloustre-Audouin ef al. (1997)
have developed a simulator of any type of landscape in 2-D in
the infrared band. Image simulators taking into account a 3-D
representation of the landscape as input exist, but they are for
* Corresponding author.
visible range or dedicated to specific applications. For example,
Thirion (1991) has developed a simulator of high spatial
resolution image in the visible range. Johnson et al. (1998) have
developed a simulator in the infrared range to simulate the
behaviour of vehicles. Guillevic (1999) was interested in
radiative budget modelling for vegetation canopy studies.
Barillot (2001) has developed MISTRAL, which can simulate
an image taking into account a 3-D description of the landscape.
A simulator of landscape adapted to remote sensing applications
does not exist and has to be developed. It should take into
account a 3-D description of the scene with high spatial
resolution.
Physical processes occurring in the signal coming from the
scene are described in the following section. Next, the
specifications of the simulator are given. The main difficulties,
related to these specifications and the infrared range are
explained. A new methodology is proposed for the simulation
of very high-resolution 3-D scenes and the consequences on the
landscape modelling are explained. Section 4 details the
architecture of the simulator using this modelling; each subpart
of the simulator is presented, and its operation is explained.
Finally, section 5 presents examples of synthesised scenes.
2. PHYSICAL PROCESSES OCCURRING IN THE
INFRARED RANGE
2.1 The radiance balance equation
In the general case, the energy equilibrium for a set of radiating
objects is expressed, independently on the wavelength A, by the
following equation:
A - 240