PCA based shadow cutting model of multiband remote sensing image

PAC based shadow cutting model fully consider the shadow´s features in the multiband remote sensing image. The shadow cutting model is based on statistical parameter of the first principal component and multiband´s combining image. First principal component after PCA tansformation contains most of the information of multiband image. Large amount of shadow information is contained in the first principal component, and is far higher than the shadow information of any single band, other principal component contains less information, in which shadows can be ignored.Using the shadow cutted image by the first principal component retransformation, this model can cut shadow effectively, and can maintain and improve the image´s quality in whole.