New Method For Target Recognition Performance Prediction Based On Triangular Pattern Meshing

Field performance prediction is a very important subject in the fields of Infrared Imaging system(IRIS) evaluation and testing. IRIS performance evaluation is based on the field performance prediction(range, detection, recognition and identification probability) or the requirements of the task. It is a very important element and the final purpose in IRIS evaluation, hence it is necessary to research how to improve the field performance prediction precision.According to TOD(Triangle Orientation Discrimination Threshold) measure idea, The recognition probability of the infrared target images at different ranges is determined focusing on typical experimental environment. The infrared target in the real background is closely meshed into many equivalent triangle patterns by considering visual acuity, and the number of triangle patterns is determined. Furthermore, the Weibull psychophysical function is used to fit the number of the equivalent triangle pattern and the corresponding recognition probability data, and the target transfer probability function is first established. Then an accurate characterization method of target thermal structure is studied, and the masking of human vision system is considered. The target recognition probability and recognition range are predicted by combining the TOD performance measure curve.