Markov-Based Image Splicing Detection in the DCT High Frequency Region

Image splicing, which is open to malicious uses, has become widespread in recent years. Therefore detection of forged images becomes a popular topic in recent years. Image splicing is one of the image forgery methods. In this study, a Markov based method in Discrete Cosine Transform domain is proposed to detect image splicing forgery. Firstly, with applying DCT to the image high frequency coefficiensts are obtained, later Markov features are extracted from the high frequency matrix. Support vector machines are used to classify images. Three commonly used data sets in the literature have been used to evaluate the performance of the method [9,10] and to compare it with similar studies. In the light of the experiments performed, authentication performance of the method gives better results when compared to others.