Dubner Computer Systems, Inc.

The Grass Valley Group, Inc.

Stuart E Lang

Jon A Fairhurst

A digital decoder for an encoded video signal has three stages. The first stage applies a multi-dimensional digital chrominance bandpass filter to the digitized encoded video signal to produce a chrominance signal, subtracting the chrominance signal from the digitized encoded video signal to produce a luminance signal. The second stage detects edges and peaks in the chrominance signal, and corrects the luminance signal where such edges and peaks are detected to produce a chrominance corrected luminance signal without dot crawl errors. An intermediate corrected chrominance signal may also be derived from the second stage. The third stage filters the chrominance signal, or intermediate corrected chrominance signal, to reduce peaks in the chrominance signal as well as to enhance edges to produce a corrected chrominance signal with minimized cross color errors.

Method of encoded video decoding

Digital prefiltering of luminance and chrominance components for an encoded video signal is performed to prevent cross color and dot crawl errors when the encoded video signal is decoded. The luminance component is filtered using a two- or three-dimensional chrominance bandpass filter to generate luminance correction limiting values corresponding to cross color errors that would occur when a conventionally encoded video signal is decoded. The luminance component also is filtered individually in each dimension, horizontally, vertically and/or temporally, to obtain luminance dimensional correction values. From the luminance correction limiting values and corresponding luminance dimensional correction values, luminance correction values are obtained by a two step decision process. The luminance correction values are added to the luminance component to provide the prefiltered luminance component. In like manner chrominance correction limiting values and chrominance dimensional correction values are generated and used to obtain chrominance correction values that are added to the chrominance component to provide the prefiltered chrominance component. The prefiltered components are combined, along with synchronizing signals, to produce the encoded video signal.

Digital prefiltering of encoded video signals

A diagonal correction method of a composite video decoder for reducing cross color errors detects the presence of diagonal luminance information in a luminance signal recovered from an encoded video signal. The recovered luminance signal is filtered to produce a diagonal correction value which is added to the recovered luminance signal as well as to a corresponding recovered chrominance signal when the diagonal luminance information is detected. The resulting luminance and chrominance signals have enhanced contrast and reduced chrominance error.

Diagonal correction in composite video decoder

A field motion suppression technique in interlaced video displays produces a motion suppressed frame without the use of a fixed threshold for motion detection. One of two interlaced fields is selected to remain unchanged, and the other field is examined on a pixel by pixel basis using vertical pixels both above and below from both fields to determine motion for each pixel. When motion is detected for a pixel, the pixel value is replaced with an interpolated value from the unchanged field, otherwise the pixel value is left unchanged.

Field motion suppression in interlaced video displays

An apparatus and method for decoding an encoded video signal combs a two dimensional array of pixels from a single field of the encoded video signal according to selected weighting factors to obtain a weighted average chrominance value. The weighted average chrominance value is subtracted from the encoded value to obtain an initial luminance value. The initial luminance value is corrected for residual chrominance errors that exist at edges in the picture represented by the encoded video signal. The luminance correction is based upon the chorminance difference across the edge and upon the orientation of the edge. The corrected luminance value may then be subtracted from the corresponding encoded video value to obtain a corrected chrominance value.

Growing community of inventors

Montville, NJ, United States of America

Stuart E Lang