针对雾天降质图像，提出了一种新的有雾偏振图像复原方法。对大气光强及传输率函数进行估计，并优化传输率函数，利用大气散射模型，对3 个任意角度的偏振图像进行图像复原，以实现图像去雾。实验比较了3 个任意角度与2 个正交角度偏振图像复原质量。实验结果表明，该方法提高了图像的清晰度及对比度，有效改善了雾天的降质图像，且3 个偏振图像的角度选取是任意的，不需对偏振角度进行配准，有利于工程的实际应用。

For fog-degraded images, a new method of restoring polarization image is proposed. Estimate the atmospheric light intensity and transmission rate function are estimated, and the transmission rate function is optimized to realize image defogging based on atmospheric scattering model for three random angle polarization images. Polarization image restoration quality of 3 random angles and 2 orthogonal angles are compared in this experiment. Experimental results show that the proposed algorithm improves the image′s definition and contrast, and effectively improves degraded images in the haze condition. The angles of 3 polarization images are arbitrary. It is not necessary to carry on the registration of the polarization angle, and beneficial for the practical application of the engineering.