Proceedings Article | 3 March 2008
Proc. SPIE. 6621, International Symposium on Photoelectronic Detection and Imaging 2007: Photoelectronic Imaging and Detection
KEYWORDS: Target detection, Staring arrays, Thermography, Spatial frequencies, Sensors, Numerical simulations, Phase shift keying, Infrared radiation, Optical simulations, Minimum resolvable temperature difference
Due to the phasing effects, the measurements of Minimum Resolvable Temperature Difference (MRTD) for Staring
array thermal imagers often get abnormal results when the targets approaching system Nyquist frequency (f<sub>n</sub>). To
simulate the relations between MRTD values and four-bar targets' frequencies, this paper introduces the concept of best
contrast. Clearly, the MRTD results are inversely proportional to the best contrasts under optimum phases, higher
contrast corresponding to a lower MRTD. On the other hand, with the spatial frequencies increasing, the target's
opening area shrinking and leads the effective infrared eradiation decreasing, this means the MRTD results are inversely
proportional to the opening area of the target. Based on these two assumptions, and through numerical simulations, this
paper depicts the tendency chart of MRTD under optimum phases to the four-bar targets' spatial frequencies. The
tendency chart adequately explains the hump curve happens at frequencies between 0.6f<sub>n</sub> and f<sub>n</sub>. From the simulations,
the maximum of MRTD values can be predicted at the frequency of 0.89f<sub>n</sub>. The tendency chart illustrated by numerical
simulation is consistent with the MRTD results get in laboratory. While in Dynamic Minimum Resolvable Temperature
Difference (DMRTD) testing, moving the four-bar targets introduces temporal effects not present in static MRTD test.
Simulation reveals that DMRTD test can get more realistic shape of the curve between 0.6f<sub>n</sub> and f<sub>n</sub>, the characteristic
hump in the static MRTD curve between 0.6f<sub>n</sub> and f<sub>n</sub> is not seen.
