Identifying photorealistic computer graphics using second-order difference statistics

Discriminating photorealistic computer graphics from natural images is an important problem in image forensics. A new distinguishing method using second-order difference statistics is proposed in this paper. Firstly, the second-order difference signals and predicting error signals of both original and calibrated images are extracted in the HSV color space, and then the variance and kurtosis of second-order difference signals and the first four order statistics of predicting error signals are extracted to be used as distinguishing features, the Fisher linear discrimination analysis is used to construct a classifier to do the differentiating job. Experimental results show that the proposed method exhibits excellent performance for the discrimination between natural images and photorealistic computer graphics, outperforms previous proposed approaches. Moreover, it has a low computational complexity.