Research On Simulation Model Of TOD Testing Platform And Performance Measurement Method

Presently, as a performance measurement method, the Triangle Orientation Discrimination(TOD) is attracting more and more attention. Since the TOD testing costs a lot of time and energy, the result is usually affected by subjective factors. In order to resolve this problem, a Simulation Model of TOD Testing Platform is presented in this paper.At first, a Triangle Pattern Generation Model was established, which generates temperature distribution images of TOD patterns with high resolution. Then, By considering the temperature signal as the input, a simulation model for the IR starring imaging system is built, which yields simulated images of patterns of different sizes and contrasts. Furthermore, A database for the measurement system is built to store the mass of testing data. Finally, based on the Back-Propagation Network, a Triangle Direction Intelligent Discrimination Model is developed.Eye-discrimination of the output images of the model is performed, the result of which is used for intelligent discrimination based on the BP network. The Weibull psychophysical functions are used to fit the statistical data of the eye-discrimination and intelligent discrimination of the triangle patterns. The TOD curve is determined from the fitted curve. Experimental results proved that the output of the system is up to the requirement of TOD performance testing, and the intelligent direction discrimination model has a similar performance to the human eye. The system is of high efficiency and practicality. It can simplify the performance measurement of the IR imaging system and can be extended to other types of imaging system.