Forged Image Detection by Analyzing Edge, Visual Saliency and Textural Features using SVM Classifier

Abstract

Nowadays, the use of Digital images is everywhere, in the magazines, in newspapers, in hospitals, in shopping malls and all over the Internet. As the development in technology is increasing day by day, at the same time the trust in images is decreasing day by day. Most common type of Image forgery is Image composition, which is also termed by the name Image Splicing. Combination of two or more images to generate a fake image is known as Image composition. It becomes very hard to differentiate between real image and fake images because of the presence of various powerful editing software. As a result, in most of the cases, there is a need to prove whether the images are real or not. This paper describes a technique for detecting forgery of composite images using Support Vector Machine (SVM) classifier. In the state of art of work, the forged image is detected by extracting Edge and Visual Saliency features. The proposed work detects the forged image by extracting Textural features in addition with Edge and Visual Saliency features. By using True Negative (TN) rate, True Positive (TP) rate and Accuracy parameters, it is found that the proposed method gives improved efficiency when compared with the existing methods.