METHOD FOR WATERMARKING FREE VIEW VIDEO WITH BLIND WATERMARK DETECTION

摘要

PURPOSE: A method for watermarking free view video is provided to omit the need of a camera parameter or something of an actual view. CONSTITUTION: View(I) is projected to a invariant space(E10). A domain variation is applied to a 1D invariant vector(i)(E11). The invariant watermarking vector(rw) is elapsed(E12). The invariant watermark vector(w=rw-r) is calculated(E13). The invariant watermark vector(v) is obtained from a spatial domain(E14). A 1D watermark vector(w) is mapped again to the spatial display domain(E15). An output watermarking view(Iw) is created by the fusion of the view and watermark signal(W)(E16).