There are no appropriate performance characteristic parameters and computational models to quantitatively compute and predict the performance of airlight rejection utilizing polarization filtering under different conditions. Therefore, a computational model for performance of airlight rejection utilizing polarization filtering is presented. This model defines the airlight rejection ratio (ARR), integrates a model for skylight relative spectral radiant power, a solar spectral model for direct irradiance on horizontal planes, an object reflection characteristic, and other factors, and gives the ARR calculation formula whose inputs contain a sun position, an optical axis direction of a camera, atmospheric visibility, object spectral reflectance, and other parameters. Simulation and experiment of this model indicate that in sunny weather and slight haze weather, the influence respectively resulting from changes of the angular distance between the observation direction and the direction of sunlight, atmospheric visibility, and object types on the ARR gradually weakens; the change of angular distance between the observation direction and the direction of sunlight intensively affects the ARR; polarization filtering achieves similar effect of airlight rejection for most objects; the larger the ARR index is, the more effectively the image contrast is improved.