Algorithm for reconstructing infrared polarization information in infrared division-of-focal-plane systems based on superpixel Stokes vectors

The division-of-focal-plane (DoFP) infrared polarization imaging system has two major advantages: it does not increase the complexity of the optical system and does not lose the temporal resolution capability of the imaging system, thus gradually becoming a development direction of infrared polarization imaging technology. The infrared polarization information reconstruction algorithm is one of the most core components in DoFP polarization imaging processing, which directly determines the quality of polarization information reconstructed by the imaging system. To address issues such as large instantaneous field of view errors and poor accuracy of polarization information reconstruction in infrared DoFP polarization imaging systems, a reconstruction algorithm for infrared polarization information in DoFP systems based on superpixel Stokes vectors is proposed. This algorithm performs demosaicing on the superpixel Stokes vector plane with lower spatial frequency to recover the first and second Stokes vectors of pixels at each polarization angle, and finally obtains complete intensity and linear polarization information for all pixels. Compared with traditional algorithms, this algorithm has the advantages of being less prone to reconstructing false polarization information, higher reconstruction accuracy, and lower latency in embedded hardware implementation, making it easy to popularize and apply.