US5168375A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Image reconstruction by use of discrete cosine and related transforms
Assignee: Intellectual Ventures I LLC
•
Filing: 1991-09-18
•
Granted: 1992-12-01
•
Expires: 2011-09-18
Abstract
Abstract A method for processing a field of image data samples to provide for one or more of the functions of decimation, interpolation, and sharpening is accomplished by use of an array transform processor such as that employed in a JPEG compression system. Blocks of data samples are transformed by the discrete even cosine transform (DECT) in both the decimation and interpolation processes, after which the number of frequency terms is altered. In the case of decimation, the number of frequency terms is reduced, this being followed by inverse transformation to produce a reduced-size matrix of sample points representing the original block of data. In the case of interpolation, additional frequency components of zero value are inserted into the array of frequency components after which inverse transformation produces an enlarged data sampling set without an increase in spectral bandwidth. In the case of sharpening, accomplished by a convolution or filtering operation involving multiplication of transforms of data and filter kernel in the frequency domain, there is provided an inverse transformation resulting in a set of blocks of processed data samples. The blocks are overlapped followed by a savings of designated samples, and a discarding of excess samples from regions of overlap. The spatial representation of the kernel is modified by reduction of the number of components, for a linear-phase filter, and zero-padded to equal the number of samples of a data block, this being followed by forming the discrete odd cosine transform (DOCT) of the padded kernel matrix.
Inventors
- 👤 Michael L. Reisch
- 👤 Munib A. Wober
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-26 06:02
CPC Classifications
G; G06; G06F; G06F17/00; G06F17/10; G06F17/14; G06F17/145; G06F17/17; G06F17/175; H; H04; H04N; H04N19/00; H04N19/10; H04N19/102; H04N19/124; H04N19/126; H04N19/132; H04N19/169; H04N19/17; H04N19/176; H04N19/18; H04N19/48; H04N19/50; H04N19/59; H04N19/85; H04N19/86; H04N19/90
Claims Analysis
Total Claims: 25
Claim 1
Independent Claim
172 words
A method for processing a field of image data samples by means of at least one array transform processor, said at least one array transform processor being operative to form a transformation between a spatial domain and a frequency domain of a block of said data samples wherein said transformation is a forward transformation or an inverse transformation and wherein said transformation furthermore is a discrete cosine transformation (DCT) or a discrete sine transformation (DST) and wherein said discrete cosine transformation is a discrete even cosine transformation (DECT) or a discrete odd cosine transformation (DOCT), said method comprising steps of: extracting a block of a number of samples from said field of samples; performing said forward transformation on said block of samples to provide a matrix of a number of frequency components; altering the frequency components in said matrix to provide a modified matrix; and performing said inverse transformation on said modified matrix of frequency components to attain a set of data samples different from the samples in said block of samples.
Claim 2
Independent Claim
172 words
A method for processing a field of image data samples by means of at least one array transform processor, said at least one array transform processor being operative to form a transformation between a spatial domain and a frequency domain of a block of said data samples wherein said transformation is a forward transformation or an inverse transformation and wherein said transformation furthermore is a discrete cosine transformation (DCT) or a discrete sine transformation (DST) and wherein said discrete cosine transformation is a discrete even cosine transformation (DECT) or a discrete odd cosine transformation (DOCT), said method comprising steps of: extracting a block of a number of samples from said field of samples; performing said forward transformation on said block of samples to provide a matrix of a number of frequency components; altering the frequency components in said matrix to provide a modified matrix; and performing said inverse transformation on said modified matrix of frequency components to attain a set of data samples different from the samples in said block of samples.
Claim 3
Dependent Claim
45 words
A method according to claim 1 wherein said step of altering is accomplished by altering the number of frequency components in said matrix, the number of samples in said set of data samples being different from the number of samples in said block of samples.
Claim 4
Dependent Claim
89 words
A method according to claim 2 wherein said inverse transformation is performed by an array transform processor having an array with more locations than the number of the frequency components in said modified matrix, said method further comprising the steps of: zero-padding said modified matrix and a matrix of inverse transform coefficients to be employed in said inverse transformation prior to said step of performing said inverse transformation; and subsequent to said step of performing said inverse transformation, discarding excess zero terms to attain said set of data samples.
Claim 5
Dependent Claim
80 words
A method according to claim 2 wherein said processing of the field of image data samples provides for a decimation of the field of data samples, said step of altering the number of frequency components in said matrix of frequency components is accomplished by deletion of highest frequency components of said matrix, said forward transformation is a DECT and the number of samples in said set of samples is less than the number of samples in said block of samples.
Claim 6
Dependent Claim
89 words
A method according to claim 4 wherein said inverse transformation is performed by an array transform processor having an array with more locations than the number of the frequency components in said modified matrix, said method further comprising said steps of: zero-padding said modified matrix and a matrix of inverse transform coefficients to be employed in said inverse transformation prior to said step of performing said inverse transformation; and subsequent to said step of performing said inverse transformation, discarding excess zero terms to attain said set of data samples.
Claim 7
Dependent Claim
124 words
A method according to claim 2 wherein said processing of the field of image data samples accomplishes an interpolation of the field of image data samples, the number of samples in said set of data samples is greater than the number of samples in said block of data samples, said step of altering the number of frequency components in said matrix of frequency components is accomplished by inserting additional higher frequency components of zero amplitude into said matrix of frequency components, said inverse transformation employs a matrix of modified inverse transformation coefficients with the number of terms in the matrix of inverse transformation coefficients being equal to the number of terms in said matrix of frequency components, and said forward transformation is a DCT.
Claim 8
Dependent Claim
89 words
A method according to claim 6 wherein said inverse transformation is performed by an array transform processor having an array with more locations than the number of the frequency components in said modified matrix, said method further comprising said steps of: zero-padding said modified matrix and a matrix of inverse transform coefficients to be employed in said inverse transformation prior to said step of performing said inverse transformation; and subsequent to said step of performing said inverse transformation, discarding excess zero terms to attain said set of data samples.
Claim 9
Dependent Claim
199 words
A method according to claim 6 wherein said forward transformation is performed by an array transform processor having an array with more locations than the number of the data samples in said block, the number of data samples in said block being equal to the number of terms in a matrix of forward transform coefficients employed in said step of performing said forward transformation, said method further comprising steps of: inserting additional samples of zero amplitude to the right of and below the samples of said block to provide a matrix of data samples having a number of terms equal to the number of locations of the array transform processor; inserting additional forward transform coefficients of zero value to the right of and below the coefficient of the matrix of forward transform coefficients to provide a modified forward transform matrix having a number of coefficients equal in number to the number of the locations of said array transform processor, said step of performing said forward transformation being accomplished by use of said matrix of data samples and said modified forward transform matrix; and discarding excess frequency components of zero value resulting from the step of performing the forward transformation.
Claim 10
Dependent Claim
74 words
A method according to claim 1 further comprising steps of: providing a filter kernel in the spatial domain for processing said field of data samples; performing said forward transformation on said filter kernel to provide a matrix of a number of kernel frequency components; and wherein said step of altering is accomplished by multiplying the frequency components in the matrix of transformed data samples by the kernel frequency components in a matrix dot-product multiplication.
Claim 11
Dependent Claim
89 words
A method according to claim 9 wherein said inverse transformation is performed by an array transform processor having an array with more locations than the number of the frequency components in said modified matrix, said method further comprising said steps of: zero-padding said modified matrix and a matrix of inverse transform coefficients to be employed in said inverse transformation prior to said step of performing said inverse transformation; and subsequent to said step of performing said inverse transformation, discarding excess zero terms to attain said set of data samples.
Claim 12
Dependent Claim
25 words
A method according to claim 9 wherein said step of performing said forward transformation on said block of data samples is accomplished by the DECT.
Claim 13
Dependent Claim
23 words
A method according to claim 11 wherein said step of performing said forward transformation on the filter kernel is accomplished by the DOCT.
Claim 14
Dependent Claim
169 words
A method according to claim 9 further comprising steps of: sectioning said field of data samples into a plurality of blocks prior to performance of said step of extracting a block, said extracting step being performed sequentially for sequentially processing each of the blocks to provide a sequence of sets of said set of data samples, the data samples of said sets differing from the data samples of said blocks by virtue of processing in said multiplying step, there being an array of processed data samples in each of said sets arranged correspondingly to an array of original data samples in corresponding ones of said blocks; arranging said sets to provide an array of partially overlapping blocks of processed data samples; and saving processed data samples at predesignated locations in each set, and discarding excess processed data samples located in overlapped regions of contiguous sets to provide a second field of filtered data samples equal in number to the data samples of the first mentioned field of data samples.
Claim 15
Dependent Claim
112 words
A method according to claim 9 further comprising steps of: modifying a matrix representation of said filter kernel in the spatial domain by deletion of a portion of the kernel components from the kernel matrix wherein the filter kernel has a linear-phase characteristic; inserting additional kernel components of zero value to the right side and below components of the filter kernel to increase the number of kernel components to equal the number of data samples in a block of data samples, said step of inserting the additional kernel components being accomplished prior to said step of performing the forward transformation of the filter kernel; and wherein said forward transformation is a DOCT.
Claim 16
Dependent Claim
187 words
A method according to claim 14 wherein said step of performing said forward transformation on said block of data samples is accomplished by the DECT; and further comprising steps of: sectioning said field of data samples into a plurality of blocks prior to performance of said step of extracting a block, said extracting step being performed sequentially for sequentially processing each of the blocks to provide a sequence of sets of said set of data samples, the data samples of said sets differing from the data samples of said blocks by virtue of processing in said multiplying step, there being an array of processed data samples in said set arranged correspondingly to an array of original data samples in corresponding ones of said blocks; arranging said sets to provide an array of partially overlapping blocks of processed data samples; and saving processed data samples at predesignated locations in each set, and discarding excess processed data samples located in overlapped regions of contiguous sets to provide a second field of filtered data samples equal in number to the data samples of the first mentioned field of data samples.
Claim 17
Independent Claim
146 words
A method for processing a field of image data samples by means of at least one array transform processor, being operative to form forward and inverse discrete cosine transformations between a spatial domain and a frequency domain of a block of said data samples using matrices of forward and inverse transform coefficients respectively comprising steps of: extracting a block of a number of samples from said field of samples; selectively modifying the matrix of at least one of the forward and inverse transform coefficients; performing said forward transformation on said block of samples to provide a matrix of a number of frequency components using the forward transform coefficients using the forward transform coefficients; performing said inverse transformation on said matrix of frequency components to attain a reconstructed set of data samples different from the samples in said block of samples using selectively modified inverse transform coefficients.
Claim 18
Independent Claim
146 words
A method for processing a field of image data samples by means of at least one array transform processor, being operative to form forward and inverse discrete cosine transformations between a spatial domain and a frequency domain of a block of said data samples using matrices of forward and inverse transform coefficients respectively comprising steps of: extracting a block of a number of samples from said field of samples; selectively modifying the matrix of at least one of the forward and inverse transform coefficients; performing said forward transformation on said block of samples to provide a matrix of a number of frequency components using the forward transform coefficients using the forward transform coefficients; performing said inverse transformation on said matrix of frequency components to attain a reconstructed set of data samples different from the samples in said block of samples using selectively modified inverse transform coefficients.
Claim 19
Dependent Claim
32 words
A method according to claim 16 wherein modifying the matrix of inverse transform coefficients comprises deleting selected high frequency components of said matrix of inverse transform coefficients for decimating the image data.
Claim 20
Dependent Claim
35 words
A method according to claim 17 further comprising the steps of: zero-padding said modified matrix of inverse transform coefficients to be employed in said inverse transformation prior to said step of performing said inverse transformation.
Claim 21
Dependent Claim
18 words
A method according to claim 16 wherein performing said forward transformation includes using modified matrix forward transfer coefficients.
Claim 22
Dependent Claim
20 words
A method according to claim 19 wherein modifying the forward coefficients comprises deleting selected high frequency components of said matrix.
Claim 23
Dependent Claim
32 words
A method according to claim 20 further comprising zero-padding said modified matrix of forward transform coefficients to be employed in said forward transformation prior to said step of performing said forward transformation.
Claim 24
Dependent Claim
184 words
Apparatus processing a field of image data samples by means of at least one array transform processor, being operative to form a transformation between a spatial domain and a frequency domain of a block of said data samples comprising: means for extracting a block of a number of samples from said field of samples; storage means for storing at least one of modified and unmodified matrices of forward and inverse transform coefficients, said matrices being modified for selectively enabling at least one of decimation and interpolation of the image data without filtering said data; selector means for selecting at least one of the modified and unmodified transform coefficients; and processor means responsive to the storage means and selector means performing said forward transformation on said block of samples to provide a matrix of a number of frequency components using at least one of modified and unmodified transform coefficients, said processor means for performing said inverse transformation on said modified matrix of frequency components to attain a set of data samples different from the samples in said block of samples using modified inverse transform coefficients.
Claim 25
Dependent Claim
184 words
Apparatus processing a field of image data samples by means of at least one array transform processor, being operative to form a transformation between a spatial domain and a frequency domain of a block of said data samples comprising: means for extracting a block of a number of samples from said field of samples; storage means for storing at least one of modified and unmodified matrices of forward and inverse transform coefficients, said matrices being modified for selectively enabling at least one of decimation and interpolation of the image data without filtering said data; selector means for selecting at least one of the modified and unmodified transform coefficients; and processor means responsive to the storage means and selector means performing said forward transformation on said block of samples to provide a matrix of a number of frequency components using at least one of modified and unmodified transform coefficients, said processor means for performing said inverse transformation on said modified matrix of frequency components to attain a set of data samples different from the samples in said block of samples using modified inverse transform coefficients.
References Cited (Backward Citations)
US4213150A
Cited
US4330833A
Cited
US4302775A
Cited
US4442454A
Cited
US4517599A
Cited
US4623923A
Cited
US4553165A
Cited
US4623913A
Cited
US4630307A
Cited
US4694328A
Cited
US4646152A
Cited
US4639769A
Cited
US4929381A
Cited
US4683496A
Cited
US4807033A
Cited
US4757373A
Cited
US4760605A
Cited
US4768082A
Cited
US4805031A
Cited
US4720871A
Cited
US4803548A
Cited
US4868654A
Cited
US4908698A
Cited
US4809070A
Cited
US4783840A
Cited
US4797729A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1995-12-01 | 1996-06-01 | 1996-12-01 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1999-12-01 | 2000-06-01 | 2000-12-01 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2003-12-01 | 2004-06-01 | 2004-12-01 | $7700.00 | Window Closed / Lapsed |