US4553165A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Transform processing method for reducing noise in an image
Assignee: Eastman Kodak Co
•
Filing: 1983-08-11
•
Granted: 1985-11-12
•
Expires: 2003-08-11
Abstract
Abstract An improved image processing method reduces noise in a sampled image while minimizing unintended distortion of image features. Image signals are generated representative of the light value of elements of the image. These signals are formed into signal arrays aligned to blocks of image elements. The signal arrays are transformed by a set of 4 by 4 Walsh-Hadamard functions into a corresponding set of coefficient signals. Certain of these coefficient signals represent the difference between the light value of each image element and an average light value over an image region smaller than the block being transformed. By modifying--i.e., coring or clipping--and inverting only these selected coefficient signals, artifacts related to the introduction of "false" edge-like structure are reduced in the reconstructed image. In addition, in a multi-stage processing method, the excluded coefficient signals may represent the input signals to the next stage.
Inventors
- 👤 Bryce E. Bayer
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-25 18:21
CPC Classifications
G; G06; G06T; G06T5/00; G06T5/10; G06T5/70
Claims Analysis
Total Claims: 26
Claim 1
Independent Claim
159 words
A method of processing an image for noise reduction, comprising the steps of: generating image signals representative of the light value of elements of the image; forming the image signals into signal arrays each aligned to a group of image elements; transforming each array of image signals into a set of coefficient signals corresponding to combinations of image signals representative of a smoothed light value over an image region including said group of image elements and differences in light value within said group of image elements; selecting a subset of coefficient signals that represent differences in light value between elements of the image and a smoothed light value over an image region smaller than the area defined by said group of image elements; modifying one or more of the coefficient signals in said selected subset in order to reduce noise in the processed image; and generating a processed image of reduced noise from said modified subset of coefficient signals.
Claim 2
Independent Claim
159 words
A method of processing an image for noise reduction, comprising the steps of: generating image signals representative of the light value of elements of the image; forming the image signals into signal arrays each aligned to a group of image elements; transforming each array of image signals into a set of coefficient signals corresponding to combinations of image signals representative of a smoothed light value over an image region including said group of image elements and differences in light value within said group of image elements; selecting a subset of coefficient signals that represent differences in light value between elements of the image and a smoothed light value over an image region smaller than the area defined by said group of image elements; modifying one or more of the coefficient signals in said selected subset in order to reduce noise in the processed image; and generating a processed image of reduced noise from said modified subset of coefficient signals.
Claim 3
Dependent Claim
50 words
The method as claimed in claim 1 wherein the step of selecting a subset of coefficient signals comprises first excluding another subset of coefficient signals that represent said smoothed light value over said smaller image region and then utilizing the remaining coefficient signals as said selected subset of coefficient signals.
Claim 4
Dependent Claim
25 words
The method as claimed in claim 2 wherein the step of transforming comprises transforming each array of image signals by use of the Walsh-Hadamard transform.
Claim 5
Dependent Claim
21 words
The method as claimed in claim 3 wherein the excluded subset of coefficient signals represent first order differences between image signals.
Claim 6
Dependent Claim
23 words
The method as claimed in claim 4 wherein the selected subset of coefficient signals represent second or higher order, differences between image signals.
Claim 7
Dependent Claim
44 words
The method as claimed in claim 2 further comprising performing the foregoing steps in a hierarchy of separate stages in which the image signals generated for each stage subsequent to the first comprise signals equivalent to the coefficient signals excluded from the preceding stage.
Claim 8
Dependent Claim
57 words
The method as claimed in claim 1 wherein the step of forming the image signals into arrays of signals comprises forming the image signals into arrays of signals such that each array includes at least one image signal in common with at least one other array whereby the aligned arrays correspond to overlapping groups of image elements.
Claim 9
Dependent Claim
48 words
The method as claimed in claim 3 wherein the step of transforming by use of the Walsh-Hadamard transform comprises modifying the Walsh-Hadamard transform such that at least some of the combinations of image signals constituting the coefficient signals use more than one contribution from the same image signal.
Claim 10
Independent Claim
113 words
A method of processing an image for noise reduction, comprising the steps of: generating image signals representative of the light value of elements of the image; transforming the image signals by a set of independent functions into coefficient signals corresponding to combinations of image signals representative of a summed light value and several differences in light value between image elements, including some coefficient signals representing first order difference and others representing second, or higher order, differences; modifying a set of coefficient signals representing the second or higher order differences so as to suppress noise in the processed image; and generating a processed image of reduced noise from the modified set of coefficient signals.
Claim 11
Independent Claim
113 words
A method of processing an image for noise reduction, comprising the steps of: generating image signals representative of the light value of elements of the image; transforming the image signals by a set of independent functions into coefficient signals corresponding to combinations of image signals representative of a summed light value and several differences in light value between image elements, including some coefficient signals representing first order difference and others representing second, or higher order, differences; modifying a set of coefficient signals representing the second or higher order differences so as to suppress noise in the processed image; and generating a processed image of reduced noise from the modified set of coefficient signals.
Claim 12
Dependent Claim
30 words
The method as claimed in claimed in claim 9 wherein the step of transforming comprises transforming the image signals by a set of independent functions characteristic of the Walsh-Hadamard transform.
Claim 13
Dependent Claim
50 words
The method as claimed in claim 9 further comprising performing the foregoing steps in a hierarchy of separate stages in which the image signals generated for each stage subsequent to the first comprise the equivalent of a set of the coefficient signals representing first order differences from the preceding stage.
Claim 14
Independent Claim
160 words
A method of image processing for noise reduction especially arranged to process signals derived from gradients within an image, said method comprising the steps of: generating image signals representative of the light value of elements of the image; forming the image signals into signal arrays aligned to blocks of image elements; transforming each array of image signals by a set of independent functions into a set of coefficient signals corresponding to combinations of image signals representative of image gradients, including some coefficient signals primarily representative of block-wide image gradients and others representative of combinations of semiblock-wide image gradients each locally contained within a partial area of each block; processing the coefficient signals for noise reduction by (a) deemphasizing a set of the coefficient signals representative of said block-wide image gradients and (b) modifying another set of the coefficient signals representative of said combinations of semiblock-wide image gradients; and generating a processed image of reduced noise from the processed coefficient signals.
Claim 15
Independent Claim
160 words
A method of image processing for noise reduction especially arranged to process signals derived from gradients within an image, said method comprising the steps of: generating image signals representative of the light value of elements of the image; forming the image signals into signal arrays aligned to blocks of image elements; transforming each array of image signals by a set of independent functions into a set of coefficient signals corresponding to combinations of image signals representative of image gradients, including some coefficient signals primarily representative of block-wide image gradients and others representative of combinations of semiblock-wide image gradients each locally contained within a partial area of each block; processing the coefficient signals for noise reduction by (a) deemphasizing a set of the coefficient signals representative of said block-wide image gradients and (b) modifying another set of the coefficient signals representative of said combinations of semiblock-wide image gradients; and generating a processed image of reduced noise from the processed coefficient signals.
Claim 16
Dependent Claim
43 words
The method as claimed in claim 12 further comprising performing the foregoing steps in a hierarchy of separate stages in which the image signals generated for each stage subsequent to the first are representative of said block-wide image gradients from the preceding stage.
Claim 17
Dependent Claim
40 words
The method as claimed in claim 13 wherein the step of forming the image signals into signal arrays in each stage further comprises overlapping the signal arrays in each stage so as to correspond to overlapping blocks of image elements.
Claim 18
Dependent Claim
29 words
The method as claimed in claim 12 wherein at least some of the combinations of image signals constituting the coefficient signals include multiple contributions from the same image signal.
Claim 19
Dependent Claim
34 words
The method as claimed in claim 12 wherein the step of transforming the image signals comprises transforming the image signals by means of a Walsh-Hadamard transform of a 4 by 4 or larger size.
Claim 20
Independent Claim
240 words
A method of image processing especially arranged to prevent the generation of unwanted artifacts during the processing of signals derived from gradients within an image, said method comprising the steps of: generating image signals representative of the light value of elements of the image; forming the image signals into signal arrays aligned to blocks of image elements; transforming each array of image signals by a set of independent functions characteristic of the Walsh-Hadamard transform into a set of coefficient signals corresponding to combinations of image signals representative of (a) a block-wide smoothed light value and (b) image gradients, including some coefficient signals primarily representative of block-wide image gradients and others representative of intercomparisons among image gradients within each block; removing from each set of coefficient signals the coefficient signal representative of the smoothed light value and at least some of the coefficient signals representative of block-wide image gradients; modifying at least some of the remaining coefficient signals in each set by (a) comparing these remaining signals to one or more thresholds and (b) altering certain of these signals as a function of the comparison; performing the foregoing steps in a hierarchy of separate stages in which the image signals generated for each stage subsequent to the first comprise signals equivalent to the coefficient signals removed from each set of coefficient signals in the preceding state; and generating processed image signals from the remaining coefficient signals, some of which are modified.
Claim 21
Independent Claim
240 words
A method of image processing especially arranged to prevent the generation of unwanted artifacts during the processing of signals derived from gradients within an image, said method comprising the steps of: generating image signals representative of the light value of elements of the image; forming the image signals into signal arrays aligned to blocks of image elements; transforming each array of image signals by a set of independent functions characteristic of the Walsh-Hadamard transform into a set of coefficient signals corresponding to combinations of image signals representative of (a) a block-wide smoothed light value and (b) image gradients, including some coefficient signals primarily representative of block-wide image gradients and others representative of intercomparisons among image gradients within each block; removing from each set of coefficient signals the coefficient signal representative of the smoothed light value and at least some of the coefficient signals representative of block-wide image gradients; modifying at least some of the remaining coefficient signals in each set by (a) comparing these remaining signals to one or more thresholds and (b) altering certain of these signals as a function of the comparison; performing the foregoing steps in a hierarchy of separate stages in which the image signals generated for each stage subsequent to the first comprise signals equivalent to the coefficient signals removed from each set of coefficient signals in the preceding state; and generating processed image signals from the remaining coefficient signals, some of which are modified.
Claim 22
Dependent Claim
49 words
The method as claimed in claim 17 wherein the step of forming the image signals into signal arrays is implemented such that each array includes at least one image signal in common with at least one other array whereby the aligned arrays correspond to overlapping blocks of image elements.
Claim 23
Independent Claim
120 words
A method of processing an image for noise reduction, comprising the steps of: generating image signals representative of the light value of elements of the image; forming the image signals into signal arrays each aligned to a group of image elements; transforming each array of image signals into a set of coefficient signals corresponding to combinations of image signals that represent differences in light value between elements of the image and a smoothed light value over an image region smaller than the area defined by said group of image elements; modifying one or more of the coefficient signals in order to reduce noise in the processed image; and generating a processed image of reduced noise from said modified coefficient signals.
Claim 24
Independent Claim
120 words
A method of processing an image for noise reduction, comprising the steps of: generating image signals representative of the light value of elements of the image; forming the image signals into signal arrays each aligned to a group of image elements; transforming each array of image signals into a set of coefficient signals corresponding to combinations of image signals that represent differences in light value between elements of the image and a smoothed light value over an image region smaller than the area defined by said group of image elements; modifying one or more of the coefficient signals in order to reduce noise in the processed image; and generating a processed image of reduced noise from said modified coefficient signals.
Claim 25
Dependent Claim
25 words
The method as claimed in claim 19 wherein the step of transforming comprises transforming each array of image signals by use of the Walsh-Hadamard transform.
Claim 26
Dependent Claim
20 words
The method as claimed in claim 20 wherein the coefficient isgnals represent second or higher order, differences between image signals.
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1988-11-11 | 1989-05-13 | 1989-11-12 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1992-11-11 | 1993-05-13 | 1993-11-12 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 1996-11-11 | 1997-05-13 | 1997-11-12 | $7700.00 | Window Closed / Lapsed |