期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2015
卷号：78
期号：2
出版社：Journal of Theoretical and Applied
摘要：A novel method has been proposed for the identification of shadows in satellite images. The method is based on the use of Ant Colony Optimization (ACO) for the identification of shadows in remotely sensed images. The existence of shadow regions in images has been a hindrance to image analysis and hence accurate shadow detection and removal is still a current research topic. The proposed work is a combination of the previous techniques and an object based technique. The proposed work first identifies the edges of all the objects in the scene and then each object is analyzed using ant colony optimization to determine whether it is a shadow or a foreground object. The shadow regions are detected in a finite number of steps, considering the various properties of the shadow regions.
