Weighting optimization with neural network for photo-response-non-uniformity-based source camera identification

Identifying the source camera of images is becoming increasingly important nowadays. A popular approach is to use a type of pattern noise called photo-response non-uniformity (PRNU). Despite that, the PRNU-based approach is sensitive towards scene content and image intensity. The identification is poor in areas having low or saturated intensity, or in areas with complicated texture. To solve the scene content problem, a weighting scheme that considers the reliability of image regions has been proposed in this paper. The proposed method uses an artificial neural network to determine the optimal weighting of each sub-block in images. Then the weightings are used to help determine the reliability of that region in identifying the source camera. The proposed method is tested against several state-of-art methods. The experiments show an encouraging result in terms of the ROC curve.