Classifying between computer generated and natural images: An empirical study from RAW to JPEG format

Tong Qiao, Xiangyang Luo, Hongwei Yao, Ran Shi

Abstract: Highlights•Relying on the designed wavelet filter, the calibrated noise is extracted from an inquiry image, possibly in three different types of image formats: RAW, TIFF, and JPEG.•We establish the generalized likelihood ratio test (GLRT) under the framework of hypothesis testing theory.•The multiple discriminators are investigated with the ability of classifying between natural images (operated by different stages of imaging process) and CG images.•By analyzing the performance of multiple discriminators, we expose the relationship existing between the classification and the extracted features from different stages of imaging procedure.