Abstrakt

Exposing Digital Forgery Detection by Illumination Using SVM classifier

P.Sreedevi, Dr.S.Karthik

Photographs often act as evidence for criminals cases in courts. It also plays a vital role in media such as newspapers, television, etc. But however, image forgery becomes easier nowadays as image editing software becomes available. Here we analyze the often utilized manipulation technique called splicing or image composition. Physics-based and statistical-based illuminate estimation technique is incorporated so that texture- and edge-based features are extracted to identify image forgery. Machine-learning approach is also employed; hence requires minimal human interaction.