Digital Image Forgery Detection using Artificial Neural Network and Auto Regressive Coefficients
E. S. Gopi, Nitya Lakshmanan, T. Gokul, S. KumaraGanesh, Prerak.r. Shah · 2006
Digitally forged photographs are so real that they do not leave any evidence of having been tampered with and can be indistinguishable from authentic photographs. Digitally processed image forgery makes the digital image data highly correlated. In this paper, we exploit this property by using auto regressive (AR) coefficients as the feature vector for identifying the location of digital forgery in a sample image. 300 feature vectors from different images are used to train an artificial neural network (ANN) and the ANN is tested with another 300 feature vectors. Two experiments were conducted. In experiment 1, manipulated images were used to train the ANN. In experiment 2 a database of forged images was used. Percentage of hit in identifying the digital forgery is 77.67%. in experiment 1 and 94.83% in experiment 2. The percentage of miss and the false alarm for the same is given as 22.33% and 32.33% in experiment 1 while it is 4.33% and 0% in experiment 2 respectively