期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2014
期号：ICETS
页码：326
出版社：S&S Publications
摘要：Photographs often act as evidence forcriminals cases in courts. It also plays a vital role inmedia such as newspapers, television, etc. Buthowever, image forgery becomes easier nowadays asimage editing software becomes available. Here weanalyze the often utilized manipulation techniquecalled splicing or image composition. Physics-basedand statistical-based illuminate estimation techniqueis incorporated so that texture- and edge-basedfeatures are extracted to identify image forgery.Machine-learning approach is also employed; hencerequires minimal human interaction.
关键词：illumination estimation; splicing; machine;learning approach and human interaction
