US6154493A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Compression of color images based on a 2-dimensional discrete wavelet transform yielding a perceptually lossless image
Assignee: Intel Corp
•
Filing: 1998-05-21
•
Granted: 2000-11-28
•
Expires: 2018-05-21
Abstract
Abstract What is disclosed is a method that includes splitting raw image data into a plurality of channels including color plane difference channels, and then compressing separately each of these channels using a two-dimensional discrete wavelet transform, the compression utilizing quantization, whereby the recovery of the compressed channel data yielding a perceptually lossless image. The method and apparatus operates on images directly in their Bayer pattern form. Quantization thresholds are defined for the quantizing which may vary depending upon the channel and DWT sub-band being processed.
Inventors
- 👤 Tinku Acharya
- 👤 Lina J. Karam
- 👤 Francescomaria Marino
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-09 10:43
CPC Classifications
H; H04; H04N; H04N11/00; H04N11/04; H04N1/00; H04N1/46; H04N1/64; H04N1/648; H04N19/00; H04N19/10; H04N19/169; H04N19/186; H04N19/60; H04N19/625; H04N19/63; H04N19/102
Claims Analysis
Total Claims: 35
Claim 1
Independent Claim
132 words
A method comprising: splitting raw image data into a plurality of channels which include one or more color plane difference channels, the splitting includes arranging a first color plane of the raw image data into a first color channel and a second color channel, generating a first difference channel that differences said first color channel from values associated with a second color plane of the raw image data, and generating a second difference channel that differences said second color channel from values associated with a third color plane of the raw image data, said first color plane having twice the number of associated values as values associated with either said second or third color plane; and compressing separately each of said channels using a two-dimensional discrete wavelet transform, said compression utilizing quantization.
Claim 2
Independent Claim
132 words
A method comprising: splitting raw image data into a plurality of channels which include one or more color plane difference channels, the splitting includes arranging a first color plane of the raw image data into a first color channel and a second color channel, generating a first difference channel that differences said first color channel from values associated with a second color plane of the raw image data, and generating a second difference channel that differences said second color channel from values associated with a third color plane of the raw image data, said first color plane having twice the number of associated values as values associated with either said second or third color plane; and compressing separately each of said channels using a two-dimensional discrete wavelet transform, said compression utilizing quantization.
Claim 3
Dependent Claim
13 words
A method according to claim 1 wherein said first color plane is Green.
Claim 4
Dependent Claim
13 words
A method according to claim 1 wherein said second color plane is Red.
Claim 5
Dependent Claim
13 words
A method according to claim 1 wherein said third color plane is Blue.
Claim 6
Dependent Claim
17 words
A method according to claim 1 wherein said raw image data is arranged in a Bayer pattern.
Claim 7
Dependent Claim
49 words
A method according to claim 5 wherein said values of the first color channel are located on a first row in the Bayer pattern and values of the second color channel are located on a second row in the Bayer pattern, said second row immediately succeeding said first row.
Claim 8
Dependent Claim
108 words
A method according to claim 1 wherein the compressing includes: performing a two-dimensional discrete wavelet transform (2-D DWT) on each channel, generating thereby for each channel a set of sub-bands, including LL sub-band containing salient channel information, the performing constituting a level of the 2-D DWT; if further resolution is desired, performing a 2-D DWT on the LL sub-band of each channel generated in the preceding level, generating thereby four new sub-bands and constituting a further level of the 2-D DWT; and quantizing said sub-bands of the 2-D DWT at each level, a separate quantization threshold defined for each said sub-band in each said channel at each level.
Claim 9
Dependent Claim
64 words
A method according to claim 7 wherein the performing a 2-D DWT includes: performing a one-dimensional discrete wavelet transform (1-D DWT) row-wise to generate an "L" band and "H" band therefrom; matrix transposing said "L" band and "H" band; and performing a 1-D DWT column-wise on said "L" and "H" bands to generate therefor said LL, an LH, an HL and an HH sub-band.
Claim 10
Dependent Claim
52 words
A method according to claim 7 wherein the quantizing includes: dividing each of the values in each sub-band in each channel by the corresponding quantization threshold; and rounding to the nearest integer the result of said division, said rounding yielding a quantized of the value version from said sub-band and said channel.
Claim 11
Dependent Claim
101 words
A method according to claim 1 wherein the compressing includes: performing a two-dimensional discrete wavelet transform (2-D DWT) on each channel, generating thereby for each channel a set of sub-bands including an LL sub-band containing salient image information, said transform including filtering coefficients modified by quantization thresholds, said thresholds defined for each channel in each sub-band, the performing constituting a level of the 2-D DWT; and if further resolution is desired, performing a 2-D DWT on each channel of the LL sub-band generated in the preceding level, generating thereby four new sub-bands and constituting a further level of the 2-D DWT.
Claim 12
Dependent Claim
24 words
A method according to claim 10 wherein the modification of filtering coefficients varies according to the high-pass or low-pass nature of the filtering coefficients.
Claim 13
Dependent Claim
13 words
A method according to claim 1, further comprising decompressing said compressed channel data.
Claim 14
Independent Claim
143 words
An apparatus for processing an image comprising: a channel generator configured to generate both color difference channels and color channels from said image by arranging a first color plane of the raw image data into a first color channel and second color channel, generating a first difference channel that differences said first color channel from values associated with a second color plane of the raw image data, and generating a second difference channel that differences said second color channel from values associated with a third color plane of the raw image data, said first color plane having twice the number of associated values as values associated with either said second or third color plane; and a compressor coupled to said channel generator to compress each channel generated by said channel generator using a two-dimensional discrete wavelet transform module (2-D DWT) for each channel.
Claim 15
Independent Claim
143 words
An apparatus for processing an image comprising: a channel generator configured to generate both color difference channels and color channels from said image by arranging a first color plane of the raw image data into a first color channel and second color channel, generating a first difference channel that differences said first color channel from values associated with a second color plane of the raw image data, and generating a second difference channel that differences said second color channel from values associated with a third color plane of the raw image data, said first color plane having twice the number of associated values as values associated with either said second or third color plane; and a compressor coupled to said channel generator to compress each channel generated by said channel generator using a two-dimensional discrete wavelet transform module (2-D DWT) for each channel.
Claim 16
Dependent Claim
28 words
An apparatus according to claim 13 wherein said 2-D DWT module is configured to generate a set of sub-band outputs for each channel, including a low-low (LL) sub-band.
Claim 17
Dependent Claim
49 words
An apparatus according to claim 14 wherein said compressor further comprises: a quantizer coupled to each said 2-D DWT module, said quantizer mapping the set of possible output values from each said 2-D DWT module to subset of values therefrom, said mapping varying for each sub-band in each channel.
Claim 18
Dependent Claim
100 words
An apparatus according to claim 15 wherein the compressor is further configured to perform multiple levels of the 2-D DWT, and if further resolution is desired, the compressor is capable of performing a 2-D DWT on at least one of said sub-bands of each channel generated in the preceding level, generating thereby a new set of sub-bands and constituting a further level of the 2-D DWT; and wherein the quantizer is capable of quantizing said sub-bands of the 2-D DWT at each level, using a separate quantization threshold defined for each said sub-band in each said channel at each level.
Claim 19
Dependent Claim
35 words
An apparatus according to claim 14 wherein each said 2-D DWT module utilizes quantization threshold modified filtering coefficients, said quantization threshold modification varying according to the particular sub-band, channel and nature of the filtering coefficients.
Claim 20
Dependent Claim
115 words
An apparatus according to claim 14 wherein said 2-D DWT module comprises: a first one-dimension discrete wavelet transform (1-D DWT) module configured to perform a row-wise DWT, to generate thereby an "L" and an "H" bands; a matrix transposer coupled to said first 1-D DWT module to receive said "L" and "H" bands, said transposer configured to transpose rows for columns in the channel data; and a second 1-D DWT module coupled to receive said transposed "L" and "H" band data, and operating upon said data to perform a column-wise DWT, said second 1-D DWT module configured to generate thereby a complete DWT level comprising said LL sub-band and an LH, HL, and HH sub-bands.
Claim 21
Independent Claim
148 words
A system comprising: an image processor, said processor configured to (1) split raw image data into a plurality of channels which include one or more color plane difference channels, by arranging a first color plane of the raw image data into a first color channel and second color channel, generating a first difference channel that differences said first color channel from values associated with a second color plane of the raw image data, and generating a second difference channel that differences said second color channel from values associated with a third color plane of the raw image data, said first color plane having twice the number of associated values as values associated with either said second or third color plane, and (2 compress said channels using quantization and discrete wavelet transforms; and an image memory coupled to said processor, said memory configured to store said compression channel data.
Claim 22
Independent Claim
148 words
A system comprising: an image processor, said processor configured to (1) split raw image data into a plurality of channels which include one or more color plane difference channels, by arranging a first color plane of the raw image data into a first color channel and second color channel, generating a first difference channel that differences said first color channel from values associated with a second color plane of the raw image data, and generating a second difference channel that differences said second color channel from values associated with a third color plane of the raw image data, said first color plane having twice the number of associated values as values associated with either said second or third color plane, and (2 compress said channels using quantization and discrete wavelet transforms; and an image memory coupled to said processor, said memory configured to store said compression channel data.
Claim 23
Dependent Claim
32 words
A system according to claim 19 further comprising: a computer system coupled to said image memory to receive said compressed channel data and configured to decompress said compressed channel data lossless image.
Claim 24
Independent Claim
188 words
A system comprising: an optical subsystem configured to guide light; an image sensor positioned to receive the light and in response generate pixel values representative of said light, said pixel values to be stored as raw image data; an image processor coupled to the image sensor, said processor configured to split the raw image data into a plurality of channels which include one or more color plane difference channels, by arranging a first color plane of the raw image data into a first color channel and second color channel, generating a first difference channel that differences said first color channel from values associated with a second color plane of the raw image data, and generating a second difference channel that differences said second color channel from values associated with a third color plane of the raw image data, said first color plane having twice the number of associated values as values associated with either said second or third color plane, said processor being capable of compressing said channels using quantization and discrete wavelet transforms; and an image memory coupled to said processor to store said compression channel data.
Claim 25
Independent Claim
188 words
A system comprising: an optical subsystem configured to guide light; an image sensor positioned to receive the light and in response generate pixel values representative of said light, said pixel values to be stored as raw image data; an image processor coupled to the image sensor, said processor configured to split the raw image data into a plurality of channels which include one or more color plane difference channels, by arranging a first color plane of the raw image data into a first color channel and second color channel, generating a first difference channel that differences said first color channel from values associated with a second color plane of the raw image data, and generating a second difference channel that differences said second color channel from values associated with a third color plane of the raw image data, said first color plane having twice the number of associated values as values associated with either said second or third color plane, said processor being capable of compressing said channels using quantization and discrete wavelet transforms; and an image memory coupled to said processor to store said compression channel data.
Claim 26
Dependent Claim
16 words
A system according to claim 21, wherein said optical subsystem is part of a digital camera.
Claim 27
Dependent Claim
15 words
A system according to claim 21, wherein said optical subsystem is part of a scanner.
Claim 28
Independent Claim
82 words
A method comprising: reading compressed image data as a plurality of channels which includes one or more compressed color plane difference channels; decompressing separately each of said channels using a two-dimensional discrete wavelet transform to yield first and second difference channels of said decompressed image; and summing the first and second difference channels with values taken from separate color planes of the decompressed image to yield first and second color channels that together completely define another color plane of the decompressed image.
Claim 29
Independent Claim
82 words
A method comprising: reading compressed image data as a plurality of channels which includes one or more compressed color plane difference channels; decompressing separately each of said channels using a two-dimensional discrete wavelet transform to yield first and second difference channels of said decompressed image; and summing the first and second difference channels with values taken from separate color planes of the decompressed image to yield first and second color channels that together completely define another color plane of the decompressed image.
Claim 30
Dependent Claim
39 words
A method according to claim 24, wherein decompressing includes: decoding said compressed image data channel by channel; inverse quantizing said decoded data; and performing an inverse 2-dimensional DWT upon said inverse quantized data, the result being recovered channel data.
Claim 31
Dependent Claim
22 words
A method according to claim 25 wherein said recovered channel data is added selectively together to recover raw image color plane data.
Claim 32
Dependent Claim
150 words
An article comprising: a machine-readable medium containing instructions which when executed by a machine cause the machine to perform operations comprising splitting raw image data into a plurality of channels which includes one or more color plane difference channels, the splitting includes arranging a first color plane of the raw image data into a first color channel and a second color channel, generating a first difference channel that differences said first color channel from values associated with a second color plane of the raw image data, and generating a second difference channel that differences said second color channel from values associated with a third color plane of the raw image data, said first color plane having twice the number of associated values as values associated with either said second or third color plane, and compressing separately each of said channels using a two-dimensional discrete wavelet transform, said compression utilizing quantization.
Claim 33
Dependent Claim
150 words
An article comprising: a machine-readable medium containing instructions which when executed by a machine cause the machine to perform operations comprising splitting raw image data into a plurality of channels which includes one or more color plane difference channels, the splitting includes arranging a first color plane of the raw image data into a first color channel and a second color channel, generating a first difference channel that differences said first color channel from values associated with a second color plane of the raw image data, and generating a second difference channel that differences said second color channel from values associated with a third color plane of the raw image data, said first color plane having twice the number of associated values as values associated with either said second or third color plane, and compressing separately each of said channels using a two-dimensional discrete wavelet transform, said compression utilizing quantization.
Claim 34
Dependent Claim
100 words
An article comprising: a machine-readable medium containing instructions which when executed by a machine cause the machine to perform operations comprising reading compressed image data as a plurality of channels which includes one or more compressed color plane difference channels; decompressing separately each of said channels using a two-dimensional discrete wavelet transform to yield first and second difference channels of said decompressed image; and summing the first and second difference channels with values taken from separate color planes of the decompressed image to yield first and second color channels that together completely define another color plane of the decompressed image.
Claim 35
Dependent Claim
100 words
An article comprising: a machine-readable medium containing instructions which when executed by a machine cause the machine to perform operations comprising reading compressed image data as a plurality of channels which includes one or more compressed color plane difference channels; decompressing separately each of said channels using a two-dimensional discrete wavelet transform to yield first and second difference channels of said decompressed image; and summing the first and second difference channels with values taken from separate color planes of the decompressed image to yield first and second color channels that together completely define another color plane of the decompressed image.
References Cited (Backward Citations)
US5541653A
Cited
US5495292A
Cited
US5412427A
Cited
EP0682438A1
Cited
US5602589A
Cited
US5880856A
Cited
US5850482A
Cited
US5901242A
Cited
US5790705A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2003-11-28 | 2004-05-29 | 2004-11-28 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2007-11-28 | 2008-05-29 | 2008-11-28 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2011-11-28 | 2012-05-29 | 2012-11-28 | $7700.00 | Window Closed / Lapsed |