Image forgery detection using SVM classifier

Unlike text, images represent an effective communication media for humans. Today we use photographs for variety of applications. Traditionally, everyone have confidence in the content of the image. Sometimes, a picture printed in a newspaper is accepted as a certification of the truthfulness of the news. As the technology develops, we can´t trust images. Image manipulation can be done easily by using photo editing tools. As a result, we have to prove the originality of images. One of the most common forms of image modification is the image composition. Composite image is created from two or more image sources. All the different images are taken from the different devices and at different world view conditions. Digital forensics field has emerged to help restore some trust to digital images. When creating a composite image, it becomes hard to match the color of one object with reference to the other. This paper describes a technique for detecting digital image forgery using illuminant color.