Wavelet transform and channel coding for data hiding in video

This paper presents an approach for video embedding based on the discrete wavelet transform and channel coding. We address the problem of embedding logos and small images within a host video. The algorithm starts with transforming each image or frame of the video sequence into the discrete wavelet domain where their coefficients are grouped into vectors equivalent to the codeword length of the signature data. The host video is not necessary for signature extraction. The method repeatedly merges the signature coefficients at various frames of the host video which provides frequency spread of the watermark to provide robustness against widely varying signal distortions including compression and filtering.