A Framework for Analyzing Motion Vector Entropy in H.265 Encoded Videos

In the domain of digital forensics, efficient and reliable detection of hidden information in compressed video files remains a challenge. This paper introduces a dynamic framework for unmasking concealed data in videos compressed using the H.265 codec, a prevalent standard in video compression. The proposed framework is underpinned by a detailed analysis of motion vectors' entropy, a critical component responsible for describing how objects shift between video frames. Entropy, as a measure of randomness, is predicted to contain anomalies in cases where information is hidden in the video. This paper capitalizes on these anomalies by assessing the entropy of motion vectors, seeking to identify and isolate them as potential markers of hidden data. This research is meant to be a blueprint in the world of video steganalysis, detecting hidden information in H.265 compressed videos. Expected to contribute significantly to video security and digital forensics fields, we wish to prompt further research into the application of entropy analysis across diverse data forms.