A Fast Watermarking Technology for Compressed Video

A technique for high capacity data hiding in MPEG-2 streams is presented. The objective is to maximise the payload while keeping robustness and simplicity. Ancillary data is embedded in the host signal by modulating the quantized block DCT coefficients of I frames. To achieve robustness, each information bit is embedded in more than one DCT coefficient within each intra coded block. The extraction process is blind. Thus, the presented technique is suitable for side information delivery. The scheme is less complex than a complete decoding process followed by watermarking in the pixel domain and reencoding. Selected results of computer simulations are also reported