A comparative study of different modes of perturbation for video watermarking based on motion vectors

Abstract

In this paper, we study the robustness of different grid types used for the embedding rules described in a previous work [1]. In this algorithm, we embed a mark in the uncompressed domain by disturbing slightly the motion vectors computed by an exhaustive BMA on blocks of size N*N. The embedding rule is based on a reference grid that allows to slightly displace the marked vectors in a neighborhood of there original marked position without missing the watermarking information. The grid tested are a square, a circular and an angular one. To increase the robustness, we embed the mark by generating a hierarchy of motion vectors to spread the mark on the lower level associated to block N*N with N=k*n.