The possibility of filtering light cloud cover in satellite imagery to expose objects beneath the clouds is discussed. A model of the cloud distortion process is developed and a transformation is introduced which makes the signal and noise additive so that optimum linear filtering techniques can be applied. This homomorphic filtering can be done in the two dimensional image plane, or it can be extended to include the spectral dimension on multispectral data. This three dimensional filter is especially promising because clouds tend to follow a common spectral response. The noise statistics can either be estimated from a general cloud model or they can be derived from a multispectral classification program. Results from a computer simulation and from LANDSAT data are shown.
Date of this Version