An Efficient Packet-level JPEG Forensic Data Collection

Recently, multimedia transmission in the Internet is popular. Digital files are widespread. However, digital distribution may suffer problems including illegal duplication, piracy and pornographic inundation. To avoid disseminating unlawful JPEG images and to control the evidences of illegal distribution are important for digital forensic. In this paper, a novel digital feature retrieval mechanism for JPEG image files is proposed. The scheme could extract JPEG binary data on the fly at the packet level using a DC feature to search for suspicious files. According to the experiment results, our method could reduce the storage volume of JPEG image files and retrieve effective features quickly from the network.