Advanced Algorithm for Shadow Detection and Reconstruction in Satellite Image Processing for Road and Forest Detection

In urban areas, the presence of shadow in high resolution satellite images is caused a serious problem for the full exploitation of images. To solve the shadow problem in high resolution satellite images, this paper propose a new shadow detection and reconstruction algorithm. Mainly three stages are using in this paper shadow detection stage, training stage and shadow reconstruction stage. The shadow detection stage consists of thresholding method, morphological filtering method and edge compensation method. The shadow reconstruction stage consists of two phases the example based learning phase and interference phase. During example based learning method ,first the pixels are sampled by manually, these samples form a shadow library and nonshadow library, which are correlated by a Markov random field(MRF).During the interference phase land cover pixels are reconstructed from the corresponding shadow pixels by adopting the Bayesian Belief Propagation algorithm to solve MRF. The algorithm is successfully tested with the datasets. QUICKBIRD, WORLDVIEW-2 and IKONOS. Index TermExample learning,Markov random field(MRF),morphological filtering ,shadow detection, shadow reconstruction.