Research On The Targeting Tasks Performance Evaluation Based On TTP Metrics

With the development of focal plane array technology, the spatial under-sampling, system magnification, background clutter have become main factors that limit the performance of thermal imaging systems. The above restrictions lead to poor performance prediction for new type of IR imager by using traditional performance based on Johnson criteria. Therefore, it is necessary to propose novel performance model or method to satisfy the performance prediction of new type of infrared imaging sensors.To solve this problem, this paper first presents NVThermIP performance prediction model based on the targeting tasks performance (TTP) criteria. Further, the impact of sampling on NVThermIP predicting performance is studied,and the spatial sampling aliasing is equivalent to be blur and noise of infrared image, respectively. Comparison experimental results show that these two models have different prediction accuracy, and the modification of sampling to NVThermIP model are implemented. Next, the effect of the system magnification and contrast enhancement on the contrast threshold function is discussed and analyzed. Finally, the contrast threshold function of infrared imaging system is established at the presence of background clutter. The relationship between the discrimination probability and range is determined.