Fast snow removal algorithm based on the maximum value of the degree of polarization and angle of polarization

Images of outdoor scenes usually have low contrast and degraded visibility in snowy weather. Snowflakes, deemed as particles in the atmosphere, cause light scattering and absorption, blurring the images. Polarization-based scenarios are proven effective in restoring the images. However, current methods either inaccurately estimate atmospheric airlight at infinity or have low efficiency. To enhance the accuracy and efficiency for increasing visibility and correcting the color shift, we propose a simple method based on polarimetric images. This method uses the maximum value of the degree of polarization and angle of polarization by global analysis of the Stokes vector to accurately estimate atmospheric airlight at infinity and the transmission map. Experimental results confirm that this method can significantly improve the image quality without complex model implementation or any other processing and has high computational efficiency. Moreover, the comparison results demonstrate that this method can recover more details and depth information.