文章基本信息

标题：Methods of Image Re-targeting Algorithm Using Markov Random Field
本地全文：下载
作者：Seyyed Mohammad Reza Hashemi ; Mohammad Mahdi Deramgozin ; Mohsen Hajighorbani 等
期刊名称：International Journal of Mechatronics, Electrical and Computer Technology
印刷版ISSN：2305-0543
出版年度：2016
卷号：6
期号：20
页码：2913-2921
出版社：Austrian E-Journals of Universal Scientific Organization
摘要：This article offers a new image re-targeting algorithm, focusing on Markov Random Field (MRF) to refine energy in image by means of Bayesian formulation. The key idea to this approach is that an energy-map is embedded into a MRF model under a Bayesian framework. The likelihood function, characterizing the saliency map likelihood at a site, is obtained based on gradient energy. These saliency maps are obtained using this likelihood function solely to model the MRF. To refine the importance maps, a priori knowledge is introduced. It will be shown that it is possible to build each of the potentials from specific PDFs. A simulated annealing algorithm is implemented to find the MAP solution. Eventually the experiments’ results, gained from the model in question, was compared to three previous methods, namely Seam Carving, , Crop, and Scaling with the performance of the image re-targeting, and the advantages of the model as well as the proposed methodology outlined at the end.
关键词：Image Retargeting; Markov Random Field; Clique; Saliency; SIFT