Advanced Video Watermarking Approach based on Convolutional Encoding : Search for new solution against cinematography piracy traffic

摘要

The massive expansion of internet network beside the permanent accessibility of online data storage resource has tremendously facilitated fraudsters task in terms of using illegally cinematic productions (Films, documentaries, animation and cartoons). In the objective to fight against movies copyright infringement and cinema piracy, we proposed in this paper, a video frame analysis based treatable system, able to keep track of unpermitted shared multimedia files. The main idea of our contribution is represented firstly by watermarking technique; each authentic multimedia file is identified by a given ID (which represent information's about the movie, for instance, time and date of projection), This ID is hidden (so as to protect purchaser privacy) using RSA encryption method, secondly, an automatic copyright checker is designated to extract the embedded ID and deduce the authenticity of the queried copy. Since multimedia files undergo video, image and audio compressions, the evoked ID is exposed to a substantial deterioration. In order to overcome this shortcoming, frame watermarking phase is reinforced by a convolutional encoding procedure. On the other hand, ID extraction is conducted via Viterbi decoding, characterized by its high efficiency in retrieving the original data in case of binary errors.