Scene classification based on rough set method features selection for outdoor images
Zeng Qing-peng, Shuixiu Wu, Mingwen Wang · 2008
Scene classification is valuable in image retrieval from databases because an understanding of the scene content can be used for efficient and effective database organization and browsing. a scene classification based on rough set method features selection for outdoor images is developed in this paper, support vector machines and k-nearest neighbors classification model are used, experiments on University of Massachusetts image segmentation database shows that based on rough set method features selection can improve the scene classification performance for outdoor images.