Visual atmospheric effects, colloquially often referred to as “haze”, are filtered from images while taking into account the directional component of such atmospheric effects. A mathematical model, taking into account directional components of light scattering causing the “haze” is utilized to provide more accurate results, especially within the context of wide field images. This mathematical model includes a directional component in the transmission coefficient. To remove the haze from images, each pixel of an individual image can have the atmospheric coloring subtracted from that pixel and can then have an appropriate compensation made for the atmospheric transmission losses. An image capture system collects metadata to aid in the determination of directional components of haze. The removal of haze from images can provide images that can be more easily combined into a composite image.