Image categorization based on visual saliency and Bag-of-Words model

摘要

Large-scale image categorization is a challenging task. In this paper, we propose a new image categorization approachbased on visual saliency and bag-of-words model. Firstly, a saliency map is generated by visual saliency method thatexploits some coarsely localized information, i.e. the salient region shape and contour. Secondly, size of salient region isacquired by calculating maximum entropy. Thirdly, the local image descriptor-SIFT extracted in the salient region andvisual saliency information are combined to build visual words. Finally, the visual word bag is categorized by SupportVector Machine. By comparing with BOW model categorization methods, experiment results show that our methods caneffectively improve the accuracy of image categorization.