Nonuniformity Correction Algorithm for TEC-less Uncooled Infrared Imaging System

摘要

The TEC-less uncooled infrared focal plane array (UIRFPA) imaging system faces the problem of nonuniformity caused by substrate temperature drift. Based on the establishment of an ambient temperature effect function, an improved nonuniformity correction (NUC) algorithm with temperature drift compensation is proposed. The first stage is offline calibration: a templates lookup table is established experimentally, and then the fitting regression coefficients are obtained using Least Squares. The second stage is real-time correction: the original readout image is preprocessed by the fitting template to compensate for nonuniformity drift, and the total variation (TV) NUC algorithm with an adaptive iterative step size is designed to correct the residual high-frequency noise. Real infrared video test shows that the algorithm effectively compensates for the nonuniformity drift while avoiding degradation and ghosting artifacts. The corrected image quality is better than that of the traditional TV-NUC method, and the roughness is reduced by nearly 40% compared to the original image.