System assessment by image simulation requires synthetic scenarios that can be viewed by the device to be simulated. In
addition to physical modeling of the camera, a reliable modeling of scene elements is necessary. Software products for
modeling of target data in the IR should be capable of (i) predicting surface temperatures of scene elements over a long
period of time and (ii) computing sensor views of the scenario.
For such applications, FGAN-FOM acquired the software products RadTherm-IR (ThermoAnalytics Inc., Calumet,
USA) and IR-Workbench (OKTAL-SE, Toulouse, France). Inspection of the accuracy of simulation results by validation
is necessary before using these products for applications. In the first step of validation, the performance of both "thermal
solvers" was determined through comparison of the computed diurnal surface temperatures of a simple object with the
corresponding values from measurements. CUBI is a rather simple geometric object with well known material parameters
which makes it suitable for testing and validating object models in IR. It was used in this study as a test body. Comparison
of calculated and measured surface temperature values will be presented, together with the results from the
FGAN-FOM thermal object code F-TOM. In the second validation step, radiances of the simulated sensor views computed
by RadTherm-IR and IR-Workbench will be compared with radiances retrieved from the recorded sensor images
taken by the sensor that was simulated.
Strengths and weaknesses of the models RadTherm-IR, IR-Workbench and F-TOM will be discussed.