金属纳米线栅是像素刻划偏振相机的核心器件,线栅上刻划了与光电探测器像元一一对应的微偏振片阵列,该阵列具有较大的非均匀性,对成像质量有较大影响。基于此,根据纳米线栅的结构与偏振传输理论,建立纳米线栅像素级矢量传输矩阵的测量数学模型,并结合矩阵最小二乘法,推导出多次测量以拟合最优传输矩阵的方法,为偏振相机像素级非均匀性矫正提供了核心的矫正参数。然后,兼顾纳米线栅刻划方式,提出了高空间分辨率的矫正算法。通过矫正后,偏振相机非均匀性由原来的2.00%降低至0.26%。外景成像实验中,目标偏振度图像的信息熵由5.34提升至15.15。结果表明,所提算法可以有效矫正纳米线栅的非均匀性,提升了偏振图像质量。 The metal nanowire grid is the core device of the pixel reticle polarizer. The grid array is marked with the micro-polarizer array corresponding to the photodetector pixels one by one. This array has large nonuniformity and has better imaging quality Great influence. Based on this, according to the structure of nanowire grid and the theory of polarization transmission, a mathematical model for the measurement of nanowire grid pixel-level vector transfer matrix is ​​established. Combined with matrix least squares method, a method of multiple measurements to fit the optimal transfer matrix is ​​deduced. It provides the core correction parameters for polarizer pixel-level non-uniformity correction. Then, taking into account the nanowire grid scoring method, a correction algorithm of high spatial resolution is proposed. After rectification, the polarization camera non-uniformity decreased from 2.00% to 0.26%. In the field experiment, the information entropy of the target polarization image is increased from 5.34 to 15.15. The results show that the proposed algorithm can effectively correct the non-uniformity of the nanowire grid and improve the quality of the polarized image.