Shadow Detection And Removal Of Multispectral Remote Sensing Images

In recent years,with the improvement of the spectral resolution,spatial resolution and temporal resolution of the sensor,the remote sensing image has the characteristics of wide observation range,high real-time and strong data integration.At present,remote sensing has been widely used in many fields,such as ecology,geology,hydrology,environmental science,atmospheric science and so on.However,the shadow region in remote sensing images is a big problem with remote sensing image becoming more and more important to people's production and life.The existence of shadow in high resolution remote sensing images greatly enhances the difficulty of remote sensing image classification,target recognition and image registration.Traditional remote sensing image shadow detection and removal algorithm can not make good use of the characteristics of remote sensing images,which can not be accurately extracted and removed accurately.The main work of this paper is through the analysis of the characteristics of shadow in remote sensing image are put forward using the improved clustering algorithm of remote sensing image shadow detection algorithm of information fusion and dictionary reconstruction hierarchical clustering based shadow removal algorithm based on.In this paper,the results of shadow detection and shadow removal are obtained on the multi spectral remote sensing data sets:1.According to the existing shadow detection algorithms of remote sensing image shadow has a low detection rate,high false detection rate,false limitations,we propose a multispectral remote sensing image shadow detection algorithm based on information fusion which can detect the shadow areas quickly and accurately in remote sensing images.The method employs data reconstruction and multi-level processing to preserve the integrity of the interior and boundary of the extracted shadow area.2.In view of the disadvantages of the existing shadow removal algorithms,such as: poor robustness,edge burr,etc.,we propose a hierarchical clustering algorithm based sample dictionary reconstruction algorithm for shadow removal of remote sensing image.This method first extracts the same kind of objects are in shadow area and non-shadow region as different representations of the sample library construction sample dictionary,then use hierarchical clustering method to reconstruct the shadow area,combined with the dictionary to complete shadow removal.Experimental results show that the proposed method has good universality and robustness.