Distinguishing Photographic and Computer Generated Images using Local Rank Transform
Swati Bhattacharya, Manish Okade · 2019 IEEE Region 10 Symposium (TENSYMP) · 2019
In recent times, the availability of multimedia in digital format and their sharing through Internet has brought forth issues like authenticity. When light passes through a digital camera, it undergoes various processing steps such as sensor nonlinearities, compression, color filter array interpolation, white balance, auto exposure, etc. The camera fingerprints left by these steps can be used to effectively distinguish between photographic and photorealistic images. In this paper, we describe a novel technique using local rank transform to distinguish between photographic and synthetic image. The key to distinguishing between these two classes is to detect the presence of demosaicing, which is carried out using the local rank transform followed by estimating a statistical feature namely coefficient of variation which is used to train a C-SVM classifier for training and testing purposes. Good accuracy has been observed for the proposed method in comparison to other existing methods in this domain.