Patent US5010504A successfully refreshed from Google Patents.
US5010504A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Digital image noise suppression method using SVD block transform
Assignee: Eastman Kodak Co
•
Filing: 1988-02-22
•
Granted: 1991-04-23
•
Expires: 2008-02-22
Abstract
Abstract A block transform image processing method employs singular value decomposition to reduce noise without affecting texture and edge detail. A nonlinear gain function based on the measured statistics of the singular values for image noise is generated. The image is filtered to produce a detail image and a low pass filtered image. The detail image is divided into blocks and the blocks are transformed to produce singular vectors and arrays of singular values. The nonlinear gain function is applied to the arrays of singular values and an inverse SVD transform is applied to the modified singular values to produce a processed detail image. The processed detail image is combined with the low pass filtered image to produce a processed image having reduced noise.
Inventors
- 👤 Hsien-Che Lee
- 👤 Hui-Jung Lee
- 👤 Heemin Kwon
- 👤 Jeanine Liang
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-29 15:17
CPC Classifications
G; G06; G06T; G06T5/00; G06T5/70; G06T5/20; H; H04; H04N; H04N19/00; H04N19/50; H04N19/503; H04N19/51; H04N19/527; H04N19/85; H04N19/86; H04N5/00; H04N5/14; H04N5/21; G06T2207/00; G06T2207/20; G06T2207/20021; G06T2207/20172; G06T2207/20192
Claims Analysis
Total Claims: 9
Claim 1
Independent Claim
140 words
A method of processing an image in a digital computer for reducing noise in the image, comprising the steps of: (a) generating a nonlinear gain function based upon the measured statistics of the SVD singular values for image noise; (b) filtering the digital image to produce a detail image and a low pass filtering image; (c) dividing the detail image into blocks; (d) transforming the blocks employing an SVD transformation to produce singular vectors and arrays of singular values; (e) applying the nonlinear gain function to the arrays of singular values to produce arrays of modified singular values; (f) performing an inverse SVD on the singular vectors with modified singular values to produce blocks of processed detail image values; and (g) combining the processed image detail values with the low pass filtered image values to produce the processed digital image.
Claim 2
Independent Claim
140 words
A method of processing an image in a digital computer for reducing noise in the image, comprising the steps of: (a) generating a nonlinear gain function based upon the measured statistics of the SVD singular values for image noise; (b) filtering the digital image to produce a detail image and a low pass filtering image; (c) dividing the detail image into blocks; (d) transforming the blocks employing an SVD transformation to produce singular vectors and arrays of singular values; (e) applying the nonlinear gain function to the arrays of singular values to produce arrays of modified singular values; (f) performing an inverse SVD on the singular vectors with modified singular values to produce blocks of processed detail image values; and (g) combining the processed image detail values with the low pass filtered image values to produce the processed digital image.
Claim 3
Dependent Claim
125 words
The method of processing a digital image claimed in claim 1, wherein said step of generating a nonlinear gain function comprises the steps of: (a) producing a noise digital image having only a noise component; (b) filtering the noise digital image to produce a noise detail and a low pass filtered noise image; (c) dividing the noise detail image into a plurality of blocks; (d) performing an SVD transformation on the blocks of the noise detail image to produce singular vectors and an array of singular values for each block; (e) calculating the means and standard deviations for respective singular values of the blocks; and (f) generating a nonlinear gain function for each of the singular values based upon the respective means and standard deviations.
Claim 4
Dependent Claim
66 words
The method of processing a digital image claimed in claim 1, further including the steps of: (a) operating the method in a plurality of stages, wherein each stage employs blocks overlapping with blocks of another stage; and (b) generating the processed digital image from the average values of the processed image values from the overlapping blocks, whereby the processed image is generated without visible block structure.
Claim 5
Dependent Claim
63 words
The method of processing a digital image claimed in claim 1, further including the step of: operating the method in a hierarchy of stages, wherein each stage employs an image detail signal representing a different pass band of spatial frequencies, and generating the process digital image by combining the processed sources characterized by different spatial frequency content is effectively removed from the image.
Claim 6
Dependent Claim
67 words
The method of processing a digital image claimed in claim 4, further including the steps of: (a) operating the method in a plurality of stages, wherein each stage employs blocks overlapping with blocks of another stage; and (b) generating the processed digital image from the average values of the processed image values from the overlapping blocks, whereby the processed image is generated without a visible block-like structure.
Claim 7
Dependent Claim
51 words
The method of processing a digital image claimed in claim 1, further including the steps of: (a) dividing the detail image into blocks having diagonally oriented edges; (b) performing the SVD transform on the diagonally oriented blocks; and (c) employing the blocks having the highest singular values for processing the image.
Claim 8
Dependent Claim
41 words
The method of processing a digital image claimed in claim 1, wherein the digital image is a color digital image, and wherein the method is applied to each color component of the digital image to produce a processed color digital image.
Claim 9
Dependent Claim
47 words
The method of processing a digital image claimed in claim 1, wherein the image is a color digital image having a luminance component and two color components, wherein the method is applied to the luminance component of the digital image to produce a processed color digital image.
References Cited (Backward Citations)
US4055756A
Cited
US4179709A
Cited
US4246610A
Cited
US4242705A
Cited
US4590608A
Cited
US4648120A
Cited
US4710822A
Cited
US4442454A
Cited
US4549212A
Cited
US4553165A
Cited
US4817180A
Cited
US4672437A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1994-04-22 | 1994-10-22 | 1995-04-23 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1998-04-22 | 1998-10-22 | 1999-04-23 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2002-04-22 | 2002-10-22 | 2003-04-23 | $7700.00 | Window Closed / Lapsed |