Abstract: Considering the advantages of the zero-watermarking algorithm and the shortage of the video watermarking for 3-dimentional high efficiency video coding. A video zero-watermarking algorithm is proposed. Firstly, the depth map of I frame and the motion vectors and coding unit partitions of non-I frame are used to select the position for watermark construction. Secondly, apply all phase biorthogonal transform and singular value decomposition to 16 x 16 blocks to obtain the most significant bit of the maximum singular value, and define the most significant bit as the final charac-teristic information. Finally, the XOR operation between the characteristic information and image watermark is per-formed to generate the zero watermark for registration. The experimental results show that the proposed algorithm has excellent robustness under different quantization parameters which can resist recompression coding attack, hybrid attacks of frame-based and recompression coding and hybrid attacks of signal processing and recompression coding.