A vision-based fully automated approach to robust image cropping detection

The definition of valid and robust methodologies for assessing the authenticity of digital information is nowadays critical to contrast social manipulation through the media. A key research topic in multimedia forensics is the development of methods for detecting tampered content in large image collections without any human intervention. This paper introduces AMARCORD (Automatic Manhattan-scene AsymmetRically CrOpped imageRy Detector), a fully automated detector for exposing evidences of asymmetrical image cropping on Manhattan-World scenes. The proposed solution estimates and exploits the camera principal point, i.e., a physical feature extracted directly from the image content that is quite insensitive to image processing operations, such as compression and resizing, typical of social media platforms. Robust computer vision techniques are employed throughout, so as to cope with large sources of noise in the data and improve detection performance. The method leverages a novel metric based on robust statistics, and is also capable to decide autonomously whether the image at hand is tractable or not. The results of an extensive experimental evaluation covering several cropping scenarios demonstrate the effectiveness and robustness of our approach.