Fraunhofer Institute of Optronics, System Technologies and Image Exploitation IOSB

TRM4 and image-based simulations

Tools for the development and performance assessment of modern optronic imagers

 

Fig. 1: Result representation of TRM4.v2 in its graphical user interface

 

Development and application of modern optronic imagers

A number of factors must be taken into account during the development and application of modern optronic imagers. The performance that can be achieved depends not only on the characteristics of the camera but also on the atmospheric conditions as well as the observed object.

Often, we need to know the "range performance" of a system with regard to a particular task. This range performance is the maximum target-observer distance at which the observer can solve the task with a specified success probability.

For an imager that is still in the developmental phase, the performance can not be determined by laboratory or field experiments. Thus, it is crucial to have different ways of assessing the imager performance as the achievable performance plays an important role in selecting suitable imager components. More generally, it is often desirable to avoid expensive and complex field experiments.

 

Task

Aside from experiments, analytical models and/or image-based simulations can be used to gain a great deal of optronic systems performance data.

Analytical models, such as TRM4, have the great advantage of delivering fast results for a number of possible system configurations and environmental conditions, but require a comparatively abstract description of the observed objects for modelling the human perception process.

Image-based simulations on the other hand emulate the imaging process and offer greater realism with regard to observation scenarios but are therefore more specific and less easy to generalize. In addition, the performance of human observers cannot easily be modelled either. That is why the evaluation of the simulated images usually takes place in observer trials. However, images obtained by image-based simulations may serve as input data for advanced image processing algorithms such as trackers and object recognition systems. 

 

The TRM4 software package

TRM (originally Thermal Range Performance Model) has been in continuous development for several decades. Whilst first versions were limited to modelling thermal imagers, TRM4 launched in 2010 provides performance assessment of imagers from the visible up to the long-wave infrared. The user can individually specify all components of the imaging process from the object through atmosphere to imager and display right up to the observer. The program contains a library of generic, spectrally resolved radiation sources, object and background reflectivities and emissivities, and detector sensitivity data.

Range performance calculation in TRM4 is based on modified Johnson criteria. The central idea of this approach is to replace actual objects by equivalent bar patterns, as shown in Fig. 2, and to find the maximum range at which an observer can resolve the equivalent bar pattern. The size, number of line pairs, and signal difference of the equivalent bar pattern depends on the object and the observation task (typically detection, recognition, and identification).

Range performance assessment of TRM4 is closely connected to the MTDP (Minimum Temperature Difference Perceived) figure of merit of thermal imagers. It is the minimum temperature difference of a standard 4-bar pattern at given spatial frequency which allows an observer to perceive the modulation of the bars. Reference [1] gives a description of the MTDP measurement. Measured MTDP values may be used as input in TRM4 as well as TRM4 can predict the MTDP of thermal imagers.

The original formulation of the TRM4 approach to imager performance assessment and an overview of the latest version, TRM4.v2, are presented in references [2,3].

Fig. 2: Substitution of an object by equivalent bar patterns (Johnson criteria)

TRM4 is one of the models that will be integrated into ECOMOS (European Computer Model for Optronic System Performance Prediction). More information about the ECOMOS project can be found at the ECOMOS website.

Apart from the TRM4 GUI version, which is distributed for free, Fraunhofer IOSB may provide customized TRM4.v2 versions that can be directly run from user applications or in batch processing mode.

 

Image-based simulations at Fraunhofer IOSB

Fraunhofer IOSB develops and applies various image-based simulations for the assessment of passive infrared imagers and active electro-optical systems. The basic idea consists in simulating the effects of different components of the electro-optical chain via a stepwise degradation of a high-quality input image, which represents the scene [4]. To a large extend the same modelling chain and the same mathematical procedures as in TRM4 can be used. Fig. 3 shows the results for such a simulation of a scanning (top right) and a staring (bottom right) thermal imager. The input was a highly resolved image, which is shown on the left side of Fig. 3.

 

Fig. 3: Image-based simulations use an input image representing the scene (left) to simulate the corresponding output images of the thermal imager. On the right hand side, output images for a scanning (top) and a staring (bottom) imager are shown.

 

 

Availability

TRM4 is available to companies and research establishments within NATO countries. Distribution and use outside NATO countries requires special permission.

 

Contact

Please contact us via trm-info@iosb.fraunhofer.de for information about TRM4 and via trm-license@iosb.fraunhofer.de for questions and requests concerning the licensing.

 

References

[1] Adomeit, U., "Measurement of the Minimum Temperature Difference Perceived (MTDP) of Thermal Imagers at Fraunhofer IOSB" [Link to the PDF]

[2] Wittenstein, W., "Minimum temperature difference perceived - a new approach to assess undersampled thermal imagers", Opt. Eng. 38, pp. 773-781, 1999 [Link to the article]

[3] Keßler, S., Gal, R. and Wittenstein, W., “TRM4: Range performance model for electro-optical imaging systems” Infrared Imaging Systems: Design, Analysis, Modeling, and Testing XXVIII, Proc. SPIE 10178, 101780P, 2017 [Link to the article]

[4] Greif, H.-J., Weiss, A. R., Wittenstein, W., "pcSitoS: a new tool for image based IR system simulation", Proc. SPIE 7481, p. 748107, 2009 [Link to the article]