A Survey of Forensics Research on Image Source Identification Forensics

Image source identification forensics is an important research direction of passive forensics of digital image, which aims to identify and detect the relevant sources of images based on the characteristics of the images themselves. Image source identification forensics mainly includes camera source identification for natural images and source analysis for non-natural images. Camera source identification mainly includes the identification of the brand, model or individual of the camera used for capturing image. Image source analysis can be divided into the detection of computer-generated (CG) images, the detection of AI synthesized images and the detection of recaptured images. In recent years, some achievements have been made in the field of image source identification forensics. Existing research methods can be roughly divided into two categories, including traditional model-based methods and deep learning-based methods. In this paper, the research methods in the field of image source identification forensics were summarized. The research framework, basic ideas, commonly used evaluation indicators and data sets were introduced. Finally, the current research status of image source identification forensics were summarized.