US5959693A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Pixel adaptive noise reduction filter for digital video
Assignee: Google Technology Holdings LLC
•
Filing: 1997-05-07
•
Granted: 1999-09-28
•
Expires: 2017-05-07
Abstract
Abstract A current pixel of a video image is adaptively filtered to provide stronger filtering when there is a high likelihood that noise is present. Weaker filtering is provided when the presence of a color and/or brightness edge is likely in the video image. Each pixel has luminance and chrominance amplitude values. In a filtering method, a current pixel is intermediate to first and second pixels in the video frame. Difference signals are determined based on the amplitude difference between the current pixel and first and second adjacent pixels. A plurality of available filters provide successively stronger filtering of the current pixel amplitude. One of the filters is selected for filtering the current pixel according to the difference signals. Secondary difference signals which indicate the amplitude difference between the current pixel and pixels which are adjacent to the first and second pixels are also considered in the selection process. The selection process may further be biased by a user-selectable control word. The scheme may also provide temporal filtering with successive video frames or images. Filtering with pixel interpolation may also be used.
Inventors
- 👤 Siu-Wai Wu
- 👤 Jay R. Patel
- 👤 Robert Nemiroff
- 👤 John Shumate
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-02 19:31
CPC Classifications
H; H04; H04N; H04N9/00; H04N9/64; H04N5/00; H04N5/14; H04N5/21
Claims Analysis
Total Claims: 30
Claim 1
Independent Claim
193 words
A method for spatially filtering pre-compressed video image data which comprises at least one video frame having a plurality of pixels to reduce the presence of noise in said data, wherein each of said pixels has an associated amplitude, and a current pixel is intermediate to first and second pixels in said video frame, comprising the steps of: determining a difference, d1, including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said first pixel; determining a difference, d2 including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said second pixel; providing a plurality of filter kernels, said filter kernels being adapted to provide successively stronger filtering of said current pixel amplitude; and selecting a particular one of said filter kernels according to said differences d1 and d2 for filtering said current pixel prior to compressing said video frame to provide stronger filtering of said current pixel amplitude when the amplitudes of said first and second pixels are both one of: (a) greater than the amplitude of said current pixel, and (b) less than the amplitude of said current pixel.
Claim 2
Independent Claim
193 words
A method for spatially filtering pre-compressed video image data which comprises at least one video frame having a plurality of pixels to reduce the presence of noise in said data, wherein each of said pixels has an associated amplitude, and a current pixel is intermediate to first and second pixels in said video frame, comprising the steps of: determining a difference, d1, including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said first pixel; determining a difference, d2 including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said second pixel; providing a plurality of filter kernels, said filter kernels being adapted to provide successively stronger filtering of said current pixel amplitude; and selecting a particular one of said filter kernels according to said differences d1 and d2 for filtering said current pixel prior to compressing said video frame to provide stronger filtering of said current pixel amplitude when the amplitudes of said first and second pixels are both one of: (a) greater than the amplitude of said current pixel, and (b) less than the amplitude of said current pixel.
Claim 3
Dependent Claim
59 words
The method of claim 1, comprising the further step of: selecting said particular filter kernel to provide stronger filtering of said current pixel amplitude when the magnitudes of said differences d1 and d2 are greater than a predetermined range which includes zero than when one of the magnitudes of said differences d1 and d2 is within said predetermined range.
Claim 4
Dependent Claim
43 words
The method of claim 1, comprising the further steps of: providing a user-selectable control signal; and selecting said particular filter kernel according to said control signal to provide one of: (a) incrementally stronger, and (b) incrementally weaker filtering of said current pixel amplitude.
Claim 5
Dependent Claim
30 words
The method of claim 1, comprising the further steps of: quantizing said differences d1 and d2; and selecting said particular filter kernel according to the quantized differences d1 and d2.
Claim 6
Dependent Claim
87 words
The method of claim 1, wherein said current, first and second pixels are intermediate to third and fourth pixels in said video frame, comprising the further steps of: determining a difference, e1, including a sign and magnitude, between the amplitude of said first pixel and the amplitude of said third pixel; determining a difference, e2, including a sign and magnitude, between the amplitude of said second pixel and the amplitude of said fourth pixel; and selecting said particular filter kernel according to said differences e1 and e2.
Claim 7
Dependent Claim
30 words
The method of claim 1, wherein: a sampling rate of pixels input to said particular filter kernel is different from a sampling rate of pixels output from said particular filter.
Claim 8
Independent Claim
222 words
A method for temporally filtering pre-compressed video image data comprising at least three successive video frames to reduce the presence of noise in said data, including a current frame, a first frame which precedes said current frame, and a second frame which follows said current frame, wherein each of said frames has a plurality of pixels, each of said pixels having an associated amplitude, and a current pixel of said current frame is intermediate to a first pixel in said first frame and a second pixel in said second frame, comprising the steps of: determining a difference, d1, including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said first pixel; determining a difference, d2 including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said second pixel; providing a plurality of filter kernels, said filter kernels being adapted to provide successively stronger levels of filtering of said current pixel amplitude; and selecting a particular one of said filter kernels according to said differences d1 and d2 for filtering said current pixel prior to compressing said current frame according to whether the amplitudes of said first, current and second pixels form a sequence which is one of: (a) monotonically increasing, (b) monotonically decreasing, (c) continuously increasing, and (d) continuously decreasing.
Claim 9
Independent Claim
222 words
A method for temporally filtering pre-compressed video image data comprising at least three successive video frames to reduce the presence of noise in said data, including a current frame, a first frame which precedes said current frame, and a second frame which follows said current frame, wherein each of said frames has a plurality of pixels, each of said pixels having an associated amplitude, and a current pixel of said current frame is intermediate to a first pixel in said first frame and a second pixel in said second frame, comprising the steps of: determining a difference, d1, including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said first pixel; determining a difference, d2 including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said second pixel; providing a plurality of filter kernels, said filter kernels being adapted to provide successively stronger levels of filtering of said current pixel amplitude; and selecting a particular one of said filter kernels according to said differences d1 and d2 for filtering said current pixel prior to compressing said current frame according to whether the amplitudes of said first, current and second pixels form a sequence which is one of: (a) monotonically increasing, (b) monotonically decreasing, (c) continuously increasing, and (d) continuously decreasing.
Claim 10
Dependent Claim
59 words
The method of claim 7, comprising the further step of: selecting said particular filter kernel to provide stronger filtering of said current pixel amplitude when the magnitudes of said differences d1 and d2 are greater than a predetermined range which includes zero than when one of the magnitudes of said differences d1 and d2 is within said predetermined range.
Claim 11
Dependent Claim
30 words
The method of claim 7, comprising the further steps of: quantizing said differences d1 and d2; and selecting said particular filter kernel according to the quantized differences d1 and d2.
Claim 12
Independent Claim
204 words
An apparatus for spatially filtering pre-compressed video image data which comprises at least one video frame having a plurality of pixels to reduce the presence of noise in said data, wherein each of said pixels has an associated amplitude, and a current pixel is intermediate to first and second pixels in said video frame, comprising: a subtractor for determining a difference, d1, including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said first pixel; a subtractor for determining a difference, d2 including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said second pixel; a plurality of filter kernels, said filter kernels being adapted to provide successively stronger levels of filtering of said current pixel amplitude; and selection means for selecting one of said filter kernels according to said differences d1 and d2 for filtering said current pixel prior to compressing said video frame to provide stronger filtering of said current pixel amplitude when the amplitudes of said first and second pixels are both greater than the amplitude of said current pixel, or the amplitudes of said first and second pixels are both less than the amplitude of said current pixel.
Claim 13
Independent Claim
204 words
An apparatus for spatially filtering pre-compressed video image data which comprises at least one video frame having a plurality of pixels to reduce the presence of noise in said data, wherein each of said pixels has an associated amplitude, and a current pixel is intermediate to first and second pixels in said video frame, comprising: a subtractor for determining a difference, d1, including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said first pixel; a subtractor for determining a difference, d2 including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said second pixel; a plurality of filter kernels, said filter kernels being adapted to provide successively stronger levels of filtering of said current pixel amplitude; and selection means for selecting one of said filter kernels according to said differences d1 and d2 for filtering said current pixel prior to compressing said video frame to provide stronger filtering of said current pixel amplitude when the amplitudes of said first and second pixels are both greater than the amplitude of said current pixel, or the amplitudes of said first and second pixels are both less than the amplitude of said current pixel.
Claim 14
Dependent Claim
57 words
The apparatus of claim 10, wherein: said selection means selects said particular kernel to provide stronger filtering of said current pixel amplitude when the magnitudes of said differences d1 and d2 are greater than a predetermined range which includes zero than when one of the magnitudes of said differences d1 and d2 is within said predetermined range.
Claim 15
Dependent Claim
46 words
The apparatus of claim 10, further comprising: means for receiving a user-selectable control signal; wherein: said selection means selects said particular filter kernel according to said control signal to provide one of: (a) incrementally stronger, and (b) and incrementally weaker filtering of said current pixel amplitude.
Claim 16
Dependent Claim
33 words
The apparatus of claim 10, further comprising: a quantizer for quantizing said differences d1 and d2; wherein: said selection means selects said particular filter kernel according to the quantized differences d1 and d2.
Claim 17
Dependent Claim
101 words
The apparatus of claim 10, wherein said current, first and second pixels are intermediate to third and fourth pixels in said video frame, further comprising: a subtractor for determining a difference, e1, including a sign and magnitude, between the amplitude of said first pixel and the amplitude of said third pixel; a subtractor for determining a difference, e2, including a sign and magnitude, between the amplitude of said second pixel and the amplitude of said fourth pixel; wherein: said selection means selects a particular one of said filter kernels according to said differences e1 and e2 for filtering said current pixel.
Claim 18
Dependent Claim
29 words
The apparatus of claim 10, wherein: a sampling rate of pixels input to said particular kernel is different from a sampling rate of pixels output from said particular filter.
Claim 19
Independent Claim
224 words
An apparatus for temporally filtering video image data comprising at least three successive video frames to reduce the presence of noise in said data, including a current frame, a first frame which precedes said current frame, and a second frame which follows said current frame, wherein each of said frames has a plurality of pixels, each of said pixels having an associated amplitude, and a current pixel of said current frame is intermediate to a first pixel in said first frame and a second pixel in said second frame, comprising: a subtractor for determining a difference, d1, including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said first pixel; a subtractor for determining a difference, d2 including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said second pixel; a plurality of filter kernels, said filter kernels being adapted to provide successively stronger levels of filtering of said current pixel amplitude; and selection means for selecting one of said filter kernels according to said differences d1 and d2 for filtering said current pixel prior to compressing said current frame according to whether the amplitudes of said first, current and second pixels form a sequence which is one of: (a) monotonically increasing, (b) monotonically decreasing, (c) continuously increasing, and (d) continuously decreasing.
Claim 20
Independent Claim
224 words
An apparatus for temporally filtering video image data comprising at least three successive video frames to reduce the presence of noise in said data, including a current frame, a first frame which precedes said current frame, and a second frame which follows said current frame, wherein each of said frames has a plurality of pixels, each of said pixels having an associated amplitude, and a current pixel of said current frame is intermediate to a first pixel in said first frame and a second pixel in said second frame, comprising: a subtractor for determining a difference, d1, including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said first pixel; a subtractor for determining a difference, d2 including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said second pixel; a plurality of filter kernels, said filter kernels being adapted to provide successively stronger levels of filtering of said current pixel amplitude; and selection means for selecting one of said filter kernels according to said differences d1 and d2 for filtering said current pixel prior to compressing said current frame according to whether the amplitudes of said first, current and second pixels form a sequence which is one of: (a) monotonically increasing, (b) monotonically decreasing, (c) continuously increasing, and (d) continuously decreasing.
Claim 21
Dependent Claim
58 words
The apparatus of claim 16, wherein: said selection means selects said particular filter kernel to provide stronger filtering of said current pixel amplitude when the magnitudes of said differences d1 and d2 are greater than a predetermined range which includes zero than when one of the magnitudes of said differences d1 and d2 is within said predetermined range.
Claim 22
Dependent Claim
33 words
The apparatus of claim 16, further comprising: a quantizer for quantizing said differences d1 and d2; wherein: said selection means selects said particular filter kernel according to the quantized differences d1 and d2.
Claim 23
Dependent Claim
20 words
The method of claim 1, wherein: each of said filter kernels provides a non-zero attenuation of said current pixel amplitude.
Claim 24
Dependent Claim
20 words
The method of claim 7, wherein: each of said filter kernels provides a non-zero attenuation of said current pixel amplitude.
Claim 25
Dependent Claim
20 words
The apparatus of claim 10, wherein: each of said filter kernels provides a non-zero attenuation of said current pixel amplitude.
Claim 26
Dependent Claim
20 words
The apparatus of claim 16, wherein: each of said filter kernels provides a non-zero attenuation of said current pixel amplitude.
Claim 27
Independent Claim
239 words
An apparatus for temporally filtering video image data comprising at least three successive video frames to reduce the presence of noise in said data, including a current frame, a first frame which precedes said current frame, and a second frame which follows said current frame, wherein each of said frames has a plurality of pixels, each of said pixels having an associated amplitude, and a current pixel of said current frame is intermediate to a first pixel in said first frame and a second pixel in said second frame, comprising: a subtractor for determining a difference, d1, including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said first pixel; a subtractor for determining a difference, d2 including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said second pixel; a plurality of filter kernels, said filter kernels being adapted to provide successively stronger levels of filtering of said current pixel amplitude; and selection means for selecting one of said filter kernels according to said differences d1 and d2 for filtering said current pixel prior to compressing said current frame to provide stronger filtering of said current pixel amplitude when the amplitudes of said first and second pixels are both greater than the amplitude of said current pixel, or the amplitudes of said first and second pixels are both less than the amplitude of said current pixel.
Claim 28
Independent Claim
239 words
An apparatus for temporally filtering video image data comprising at least three successive video frames to reduce the presence of noise in said data, including a current frame, a first frame which precedes said current frame, and a second frame which follows said current frame, wherein each of said frames has a plurality of pixels, each of said pixels having an associated amplitude, and a current pixel of said current frame is intermediate to a first pixel in said first frame and a second pixel in said second frame, comprising: a subtractor for determining a difference, d1, including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said first pixel; a subtractor for determining a difference, d2 including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said second pixel; a plurality of filter kernels, said filter kernels being adapted to provide successively stronger levels of filtering of said current pixel amplitude; and selection means for selecting one of said filter kernels according to said differences d1 and d2 for filtering said current pixel prior to compressing said current frame to provide stronger filtering of said current pixel amplitude when the amplitudes of said first and second pixels are both greater than the amplitude of said current pixel, or the amplitudes of said first and second pixels are both less than the amplitude of said current pixel.
Claim 29
Independent Claim
237 words
A method for temporally filtering pre-compressed video image data comprising at least three successive video frames to reduce the presence of noise in said data, including a current frame, a first frame which precedes said current frame, and a second frame which follows said current frame, wherein each of said frames has a plurality of pixels, each of, said pixels having an associated amplitude, and a current pixel of said current frame is intermediate to a first pixel in said first frame and a second pixel in said second frame, comprising the steps of: determining a difference, d1, including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said first pixel; determining a difference, d2 including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said second pixel; providing a plurality of filter kernels, said filter kernels being adapted to provide successively stronger levels of filtering of said current pixel amplitude; and selecting a particular one of said filter kernels according to said differences d1 and d2 for filtering said current pixel prior to compressing said current frame to provide stronger filtering of said current pixel amplitude when the amplitudes of said first and second pixels are both greater than the amplitude of said current pixel, or the amplitudes of said first and second pixels are both less than the amplitude of said current pixel.
Claim 30
Independent Claim
237 words
A method for temporally filtering pre-compressed video image data comprising at least three successive video frames to reduce the presence of noise in said data, including a current frame, a first frame which precedes said current frame, and a second frame which follows said current frame, wherein each of said frames has a plurality of pixels, each of, said pixels having an associated amplitude, and a current pixel of said current frame is intermediate to a first pixel in said first frame and a second pixel in said second frame, comprising the steps of: determining a difference, d1, including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said first pixel; determining a difference, d2 including a sign and magnitude, between the amplitude of said current pixel and the amplitude of said second pixel; providing a plurality of filter kernels, said filter kernels being adapted to provide successively stronger levels of filtering of said current pixel amplitude; and selecting a particular one of said filter kernels according to said differences d1 and d2 for filtering said current pixel prior to compressing said current frame to provide stronger filtering of said current pixel amplitude when the amplitudes of said first and second pixels are both greater than the amplitude of said current pixel, or the amplitudes of said first and second pixels are both less than the amplitude of said current pixel.
References Cited (Backward Citations)
US4107736A
Cited
US4009334A
Cited
US4240106A
Cited
US4573070A
Cited
US4249210A
Cited
US4249209A
Cited
US4275418A
Cited
US4399461A
Cited
US4278996A
Cited
US4792854A
Cited
US4571511A
Cited
US4670775A
Cited
US4672429A
Cited
US4725885A
Cited
US5023919A
Cited
US4918528A
Cited
US4783840A
Cited
US5040225A
Cited
US4926261A
Cited
US5294984A
Cited
US5025478A
Cited
US4987481A
Cited
US5025316A
Cited
US5151787A
Cited
US5602942A
Cited
US5025312A
Cited
US5321511A
Cited
US5293541A
Cited
US5363210A
Cited
US5400082A
Cited
US5392137A
Cited
US5327242A
Cited
US5491520A
Cited
US5491519A
Cited
US5512956A
Cited
US5745187A
Cited
US5442407A
Cited
US5574512A
Cited
US5606361A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2002-09-27 | 2003-03-29 | 2003-09-28 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2006-09-27 | 2007-03-29 | 2007-09-28 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2010-09-27 | 2011-03-29 | 2011-09-28 | $7700.00 | Window Closed / Lapsed |