Shadow removal based on separated illumination correction for urban aerial remote sensing images.

•A novel separated illumination correction algorithm is proposed to remove shadows for urban aerial remote sensing images.•A spatially adaptive weighted total variation model is constructed to separate the structured illumination and the shadow-free reflectance.•The object-oriented local color transfer is used to correct the shadow information in the illumination.•Shadow standard deviation index and classification analysis are introduced to evaluate the shadow removal results quantitatively.