A Review On Digital Video Watermarking Using DWT

In this paper, we propose a robust perceptual digital video watermarking procedure to embed a watermark image in digital video frames using the variable-temporal length 3-D DCT technique. A variable-length 3-D DCT is chosen for exploiting the redundancy of the video sequences in temporal domain properly. The variable window length is obtained by applying a scene-change detection to a sequence of 8 8 blocks in successive frames. The method adopted embeds the watermark (a binary logo image) in an uncompressed video sequence by modifying the values of the mid-range coefficients of individual 3-D DCT blocks, in accordance with sensitivity of the 3-D human visual system model. The sensitivity of human eye in this frequency range is known to be minimum. This preserves the perceptual quality of video sequences. Since each block consists of the both spatial as well as temporal information of the video, we could achieve to spread the watermark in both spatial and temporal domain. This makes the proposed method highly robust against common signal processing manipulations such as frame averaging, frame dropping, noise addition, cropping and lossy compression. The retrieval rate of the watermark logo image varied between 50 % to 99 % for various attacks. The lower limit of watermark extraction occurs, when half or more than half of total video frames is dropped and also in the case of cropping where percentage of area cropped is more than 45 %.