We suggest using a technology called De-haze-Net in this procedure to estimate medium transmission. De-haze-Net receives a hazy image as input and returns a medium transmission map that is then combined with an atmospheric scattering model to recover a haze-free image. De-haze-Net based deep architecture layers are specifically made to extract the characteristics of the hazy pixels, which are capable of producing practically all haze related features. Additionally, we suggest a nonlinear activation function for De-haze-Net termed a bilateral rectified linear unit, which can enhance the quality of the haze-free picture that is reconstructed. We develop links between the elements employed in conventional approaches and those proposed for the De-haze-Net.