Abstract: An encoder generates an estimate of apparent inter-image motion utilizing a dense motion field in a sequence of video frames or other images. The estimate in an illustrative embodiment includes a motion vector for each pixel of the dense motion field. The estimate is generated at least in part as a solution to a constrained optimization problem based on a multiscale data model characterizing the motion between pixels of the dense motion field and designated neighbor pixels. The estimate is used for motion compensation in encoding the corresponding images, and is itself encoded for transmission to a decoder for use in decoding the encoded images. In the illustrative embodiment, the multiscale data model characterizes spatial, temporal and scale coherence properties of the motion as a Markov random field (MRF), thereby ensuring that the resulting motion vectors can be efficiently coded for transmission to the decoder.

Abstract: Digital watermark information is inserted into an image by first separating the image into components, e.g., discrete cosine transform (DCT) blocks or image subbands, and then associating one or more bits of the digital watermark information with each of the components. For example, a single bit may be associated with each of the components by modulating the components with selected waveforms representative of the corresponding digital watermark information bits. As another example, the digital watermark information bits may be coded, e.g., using a repetition code, linear block code or convolutional code, to form channel bits, such that the modulating waveforms are selected for the image components based on the corresponding channel bits. The digital watermark information may include a total of B bits of information for representing a particular watermark, such that M=2B distinct watermarks can be generated using the B information bits.