Abstract: An input data stream of pixel values, representing an original video image having predetermined horizontal and vertical resolutions that has been spatially sampled at a first spatial sampling frequency, is upsampled and interpolated to derive an output data stream of interpolated pixel values at a second spatial sampling frequency that represents an oversampled video image that has been enlarged in size with respect to the original video image but has the same full bandwidth as the original video image. The pixel values of the input data stream and the interpolated pixel values of this output data stream are both employed to synthetically derive the interpolated pixel values of a correction-waveform data stream that defines spatial frequencies that are higher in frequency than the highest frequency in the full bandwidth of the original video image.

Abstract: A compressed data system, which communicates content of each of successive original frames of a digital motion picture over a limited-bandwidth transmission channel, comprises a digitally-controlled pre-processor and a digitally-controlled post-processor. The digitally-controlled pre-processor is effective in reducing the spatial-frequency bandwidth of the content of an original frame, applied to the input of a compression encoder, in response to the digital control value of a first digital control signal, applied as a control input to the pre-processor, being indicative of the fact that the bandwidth of the compressed data to be communicated over the limited-bandwidth transmission channel would exceed the limited bandwidth of the transmission channel. Further, the pre-processor derives one or more digital control values of a second digital control signal communicated over the limited-bandwidth transmission channel without ever being compressed directly to the digitally-controlled post-processor.

Abstract: Filtering apparatus responsive to a sequence of samples of an input signal that is digitally-sampled at a given frequency, wherein the input signal defines information having at least one dimension. The apparatus includes first resampler for resampling the input-signal to derive a first signal at a first frequency which is substantially equal to the reciprocal 1/C of a given factor C times the given frequency, and second resampler for resampling the first signal to derive a second signal at a second frequency which is substantially equal to the given factor C times the first frequency. The given factor C is an improper fraction having a least-common-denominator larger than one. This filtering apparatus is particularly suitable for variably spatially low-pass, high-pass, band-pass or hierarchical-pyramid filtering two-dimensional image information in each of its two-dimensions by amounts which are independent of one another.