Classification of image based on semantic features and Bayesian networks
Chen Hongjun, Zhang Junfeng · 2011
Traditional content-based image classifications often fail to meet a user's need due to the `semantic gap' between the texture features and the semantic features of the image. Content-based indexing and retrieval of images requires a proper semantic description for image content. This paper presents a novel approach based on semantic features and Bayesian networks for image classification. A mapping between low-level visual information and higher-level semantic space by using priori knowledge is created for image classification using Bayesian networks. We performed experiments on a set of images which are collected from web pages and simulation results show feasibility and effectiveness.