Digital image forensics with finite ridgelet transform

With the use of advanced computer graphics rendering software, very successful photorealistic images can be generated. Therefore, it may be hard to discriminate the photographic images from the photorealistic ones. In this work, we propose to use the statistics obtained from the ridgelet transform to distinguish the photographic and photorealistic images. Experimental results show that the features obtained by using the ridgelet transform is relatively less complex and more successful for the classification from the ones obtained by using wavelet transform.