Model heterogeneity correction algorithm for infrared division of focal plane polarization detectors

Infrared division-of-focal-plane (DoFP) polarization detectors often exhibit pronounced nonlinear response characteristics that cannot be effectively handled by existing nonuniformity correction algorithms. In this paper, we first establish an imaging model for infrared DoFP polarization detectors. To satisfy the requirements of infrared nonlinear DoFP polarization detectors, a model heterogeneity (MHE) nonuniformity correction algorithm is proposed. The proposed method constructs a mapping between the nonlinear response and a linear correction model, enabling effective suppression of nonuniformity noise while compensating for the correction-induced impact on polarization information. Experimental results demonstrate that the MHE algorithm significantly outperforms comparison methods in polarization preservation, nonuniformity noise suppression, and real-image correction. Compared with the comparison algorithms, the RMSE is substantially reduced. Meanwhile, the PCM is reduced from 70.1-278.4 to 14.8, while the APMR is improved from 16.9-22.8 to 33.9.