Re-compressed based JPEG forgery detection and localization through automated quality factor investigation

摘要

In today's media-saturated society, digital images act as the primary carrier for majority of information that flows around us. However, because of the advent of highly sophisticated easy-to-use image processing tools, modifying images has become easy. In this paper we deal with detection of forgeries in Joint Photographic Experts Group (JPEG) images. Due to its best compression features and optimal space requirement, JPEG is used as a world-wide standard format for storing digital images. Tampering in a JPEG images often involve re-compression. In this paper we proposed a forensic technique for automatic detecting and localizing of such re-compression based JPEG forged regions. The proposed technique is based on estimating the quality factor of the tampered image block-wise, in order to detect and localize the forgery. The proposed technique is capable of detecting single as well as multiple forged regions in a JPEG image.