Faroudja Enterprises, Inc.

Other

Genesis Microchip Corporation

Sage, Inc.

Xu Dong

Yves C Faroudja

Jack J Campbell

Ka Lun Choi

Donald S Butler

Ravi Ananthapur Bacche

Daniel Campbell

Minqiang Jiang

Dong Wei

A novel de-blocking and de-banding filter and a related method of operation reduce and/or remove blocking and banding artifacts that originate from a low bitrate video compression. For filtering a luminance channel, the de-blocking and de-banding filter determines filtering needs and also adjusts filtering strengths based on a compression level (Q) and a gray level of the video signal. For filtering a chrominance channel, the de-blocking and de-banding filter determines the filtering needs and adjusts the filtering strengths based on Qalone. Preferably, the de-blocking and de-banding filter utilizes a low pass filter of adjustable strength at a border of a block perpendicular to the border or a band border. The adjustable strength of the de-blocking and de-banding filter is a function of the compression level (Q) and/or the gray level of the video signal, but does not depend on a duration of the blocking and/or banding artifacts.

De-blocking and de-banding filter with adjustable filter strength for video and image processing

A video bit-rate reduction system utilizing a reference images matrix and a method of operating the video bit-rate reduction system are disclosed. In one embodiment, the video bit-rate reduction system includes an encoder-side bit-rate reduction system (e.g. a video signal transmitter) with a reference images matrix matchmaker and a decoder-side bit-rate reduction system (e.g. a video signal receiver) with a reference images matrix reconstructor. The video bit-rate reduction system is designed to accommodate a reference images matrix-based symbolization, matchmaking, and reconstruction of residual signals that are processed through a support layer pathway with a high bit-rate reduction and data transmission efficiency, while retaining the high-quality of the video data without any or substantial visible degradation of image and video quality. In a preferred embodiment, a sparse linear model is applied to the reference images matrix matchmaking and the reference images matrix reconstruction for compact symbolization of the residual signals.

Video bit-rate reduction system and method utilizing a reference images matrix

Methods and systems for bidimensional video processing comprise preprocessing the video signal in horizontal and vertical dimensions of the video signal to reduce an amount of information of the video signal by low pass filtering the video signal and removing pixels to obtain an information-reduced video signal; expanding the information-reduced video signal by an expander process to produce an upconverted video signal; and comparing the upconverted video signal to the original video input to generate a support signal based on a difference between the original video signal and the upconverted video signal. Further embodiments include at least one of storing and transmitting the information-reduced video signal and the support signal; and during post-processing, processing and combining the information-reduced video signal and the support signal to generate a full information duplicate of the video signal.

Bidimensional bit-rate reduction processing

A combiner processing system and a method of operating the combiner processing system are disclosed. In one embodiment, the combiner processing system is utilized for a support layer processing in a bit-rate reduction system that separates video data into a main layer and a support layer with separated encodings for the two layers for bit-rate reduction and data transmission efficiency. Preferably, the combiner processing system for the support layer processing is capable of processing repeat patterns, small high frequency details, and textual information contained in one or more images. Furthermore, in one embodiment, the combiner processing system for the support layer processing provides a motion vector detection on a full image reconstruction, a non-linear quantization which retains both textual information and small high frequency details, utilization of full combination motion vectors, and a separate processing path for textual information with an internal bit-rate reduction unit.

Combiner processing system and method for support layer processing in a bit-rate reduction system

One or more systems and methods for edge-adaptive and recursive non-linear filtering of ringing effect on image or video data are disclosed in accordance with various embodiments of the invention. In one embodiment of the invention, a method for edge-adaptive and recursive non-linear filtering of ringing effect first involves detecting all edges within a process area of an image, including a direction of an edge slope and a value of an edge signal level. Then, if the process area is determined to be a 'non-busy' area based on a 'busyness' measure relative to empirically-defined threshold values, then an edge influence function subsequently determines whether to apply a de-ringing filter to a current pixel within the process area or not. Preferably, the de-ringing filter is edge-adaptive, non-linear, and recursive. The de-ringing filter can adjust the current pixel multiple times based on filter angles, adjacent pixels, and pixel transition levels.

System and method for edge-adaptive and recursive non-linear filtering of ringing effect

Methods and apparatus, including computer program products, implementing and using techniques for distinguishing false motion from true motion, in a video signal that is received by a television set. The video signal is received at the television set. The video signal includes a luma component and a chroma component. A motion detection operation is performed for each of the chroma component and the luma component in several frequency regions of the received video signal. Each frequency region has a unique, adjustable threshold value for distinguishing false motion from true motion, that is dynamically adjustable based on the incoming video signal, and as a result the motion detection accuracy is improved and image artifacts are minimized in an image that is described by the video signal and is to be displayed on the television set.

Motion detection in video signals

An interlaced television signal is derived from an interlaced 625 line, nominally 50 Hz field rate television signal, the derived television signal having perceived reduced line structure and reduce flicker. The field rate and the number of lines of the derived television signal are increased with respect to the field rate and the number of lines of the original television signal, such that perceived flicker and line structure in the derived television signal is reduced. The increase in the field rate and the increase in the number of lines in the derived television signal results in a horizontal scanning rate that does not substantially exceed twice the horizontal scanning rate of the original television signal while minimizing undesirable motion artifacts.

Reduced television display flicker and perceived line structure with low horizontal scan rates

An interlaced television signal is derived from an interlaced 625 line, nominally 50 Hz field rate television signal, the derived television signal having perceived reduced line structure and reduced flicker. The field rate and the number of lines of the derived television signal are increased with respect to the field rate and the number of lines of the original television signal, such that perceived flicker and line structure in the derived television signal is reduced. The increase in the field rate and the increase in the number of lines in the derived television signal results in a horizontal scanning rate that does not substantially exceed twice the horizontal scanning rate of the original television signal while minimizing undesirable motion artifacts.

A convectional 2-1 interlaced (nominally 50 Hz field rate or nominally 60 Hz field rate) television signal is converted to a form suitable for display on a progressively-scanned variable-frame-rate high-resolution computer-type monitor. First, a succession of progressively-scanned video frames is derived from the interlaced television signal and each frame is repeated at least twice in succession When the interlaced television sign is derived from a motion picture source, the progressively scanned frames are derived by merging opposite polarity fields derived from tho same motion picture frame. For other sources, other interlace-to-progressive scan techniques, such as line duplication or line interpolation, are employed. The sequences of repeated progressively-scanned frames are written into a two frame memory at rate derived from the timing of the progressively-scanned video frames and are read out at a rate derived from the timing of the monitor's frame rate. Frames are skipped or duplicated in order to prevent memory underflow or overflow. Motion-discontinuity-free video displays at refresh rates from 50 Hz to 100 Hz, for 50 Hz inputs, and from 60 Hz to 120 hz for 60 Hz inputs may be achieved.

Growing community of inventors

Los Altos, CA, United States of America

Xu Dong