Risk Evaluation and Detection of Information Hidden in Video Container Formats

The diversity of video file formats and the redundant space that exists in digital video file formats lead to risks from information that may be hidden in video container formats. Previous works concerning video steganalysis have focused primarily on information hidden using techniques based on the data used to encode video but have neglected information-hiding techniques based on the video container file formats themselves. In this paper, we propose a method to evaluate the risk of information being hidden in video container formats. The method can return a recommended risk level for manual steganalysis or automatic steganalysis. We then further propose two types of steganalysis methods to detect the information hidden in video container formats, including statistical pattern-based steganalysis and machine learning-based steganalysis. A case study and implementation are given for statistical pattern-based detection and machine learning-based detection, using FLV and PS video container formats as examples. The results of experiments concerning risk evaluation and these methods are given to show the effectiveness of the proposed information-hiding risk evaluation and detection methods.