Granted Patent B1

US 9,479,799 · App. 14/846,719 · Granted Oct 25, 2016

Abstract

Claims (73)

1. An apparatus for compressing an image; comprising:

a principal component analyzer configured to iteratively perform a Principal Component Analysis (PCA) on a selected portion of the image, wherein each resulting orthogonal basis set has “N” basis vectors;

a matrix rank evaluator to evaluate the rank of each orthogonal basis set;

the principal component analyzer further configured to iteratively perform the PCA based on the rank of an initial orthogonal basis set as evaluated by the matrix rank evaluator, whereby the initial orthogonal basis set exhibiting a low rank results in a negative determination with respect to the performance of a subsequent PCA and vice versa;

a puncturer configured to puncture selected dimensional components of the orthogonal basis set resulting from each PCA of the selected portion of the image, without removing any of the associated “N” basis vectors thereof;

a truncator configured to remove selected basis vectors of a final one of the resulting orthogonal basis sets of the principal component analyzer, thereby forming a truncated basis set for compression of the selected portion of the image;

a projector configured to project the image onto the truncated basis set thereby obtaining coefficients of the selected portion of the image, relative to the truncated basis set.

2. An apparatus for compressing an image; comprising:

a principal component analyzer configured to iteratively perform a Principal Component Analysis (PCA) on a selected portion of the image, wherein each resulting orthogonal basis set has “N” basis vectors;

a matrix rank evaluator to evaluate the rank of each orthogonal basis set;

the principal component analyzer further configured to conditionally iteratively perform the PCA based on the rank of each orthogonal basis set as evaluated by the matrix rank evaluator, whereby an orthogonal basis set exhibiting a low rank results in a negative determination with respect to the performance of another iteration and vice versa;

a puncturer configured to puncture selected dimensional components of the orthogonal basis set resulting from each PCA of the selected portion of the image, without removing any of the associated “N” basis vectors thereof;

a truncator configured to remove selected basis vectors of a final one of the resulting orthogonal basis sets of the principal component analyzer, thereby forming a truncated basis set for compression of the selected portion of the image;

a projector configured to project the image onto the truncated basis set thereby obtaining coefficients of the selected portion of the image, relative to the truncated basis set.

3. The apparatus for compressing an image of claim 2 , further comprising:

the principal component analyzer further configured to conditionally iteratively perform the PCA based on the rank of each orthogonal basis set as evaluated by the matrix rank evaluator, wherein the rank of a basis set corresponds with a maximal number of linearly independent rows or columns of the basis set, and whereby an orthogonal basis set exhibiting a low rank results in a negative determination with respect to the performance of another iteration and vice versa.

4. An apparatus for compressing an image; comprising:

a principal component analyzer configured to iteratively perform a Principal Component Analysis (PCA) on a selected portion of the image, wherein each resulting orthogonal basis set has “N” basis vectors;

a puncturer configured to puncture selected dimensional components of the orthogonal basis set resulting from each PCA of the selected portion of the image, without removing any of the associated “N” basis vectors thereof; and the puncturer further configured to transform each of the “N” basis vectors in the prior orthogonal basis set into a corresponding Fourier series and to puncture higher frequency ones of the Fourier series coefficients associated with the “N” basis vectors

a truncator configured to remove selected basis vectors of a final one of the resulting orthogonal basis sets of the principal component analyzer, thereby forming a truncated basis set for compression of the selected portion of the image; and

a protector configured to project the image onto the truncated basis set thereby obtaining coefficients of the selected portion of the image, relative to the truncated basis set.

5. The apparatus for compressing an image of claim 4 , further comprising:

the puncturer further configured to transform each of the “N” basis vectors in the prior orthogonal basis set into a corresponding Fourier series, to quantize a phase component of the coefficients of each of the “N” basis vector's associated Fourier series; and to puncture higher frequency ones of the Fourier series coefficients associated with the “N” basis vectors.

6. A method for compressing an image; comprising:

iteratively performing a Principal Component Analysis (PCA) on a selected portion of the image, wherein each resulting orthogonal basis set has “N” basis vectors; and with each iteration including:

puncturing of selected dimensional components of the orthogonal basis set resulting from each PCA of the selected portion of the image, without removing any of the associated “N” basis vectors thereof; and

wherein the iterative act further comprises:

performing an initial Principal Component Analysis (PCA) on the selected portion of the image, wherein the resulting “N” basis vectors form an initial orthogonal basis set;

puncturing selected dimensional components of the initial orthogonal basis set without removing any of the associated “N” basis vectors thereof; whereby the initial orthogonal basis set becomes a nearly orthogonal basis set;

performing a subsequent PCA on the nearly orthogonal basis set, thereby forming the final orthogonal basis set;

removing selected basis vectors of a final one of the resulting orthogonal basis sets of the iteration act; thereby forming a truncated basis set for compression of the selected portion of the image; and

projecting the image onto the truncated basis set thereby obtaining coefficients of the selected portion of the image, relative to the truncated basis set.

7. A method for compressing an image; comprising:

iteratively performing a Principal Component Analysis (PCA) on a selected portion of the image, wherein each resulting orthogonal basis set has “N” basis vectors; and with each iteration including:

puncturing of selected dimensional components of the orthogonal basis set resulting from each PCA of the selected portion of the image, without removing any of the associated “N” basis vectors thereof; and

wherein the iterative act further comprises:

conditionally iteratively performing the PCA based on a determination of a rank of an initial orthogonal basis set, whereby the initial orthogonal basis set exhibiting a low rank results in a negative determination with respect to the performance of the puncturing and subsequent PCA and vice versa;

removing selected basis vectors of a final one of the resulting orthogonal basis sets of the iteration act; thereby forming a truncated basis set for compression of the selected portion of the image; and

projecting the image onto the truncated basis set thereby obtaining coefficients of the selected portion of the image, relative to the truncated basis set.

8. A method for compressing an image; comprising:

iteratively performing a Principal Component Analysis (PCA) on a selected portion of the image, wherein each resulting orthogonal basis set has “N” basis vectors; and with each iteration including:

puncturing of selected dimensional components of the orthogonal basis set resulting from each PCA of the selected portion of the image, without removing any of the associated “N” basis vectors thereof; and

wherein the iterative act further comprises:

conditionally iteratively performing the PCA based on a determination of a rank of each orthogonal basis set, whereby an orthogonal basis set exhibiting a low rank results in a negative determination with respect to the performance of the puncturing and subsequent PCA and vice versa;

removing selected basis vectors of a final one of the resulting orthogonal basis sets of the iteration act; thereby forming a truncated basis set for compression of the selected portion of the image; and

projecting the image onto the truncated basis set thereby obtaining coefficients of the selected portion of the image, relative to the truncated basis set.

9. The method for compressing an image of claim 8 , wherein the iterative act further comprises:

conditionally iteratively performing the PCA based on a determination of a rank of each orthogonal basis set, wherein the rank of a basis set corresponds with a maximal number of linearly independent rows or columns of the basis set, and whereby an orthogonal basis set exhibiting a low rank results in a negative determination with respect to the performance of another iteration and vice versa.

10. A method for compressing an image; comprising:

iteratively performing a Principal Component Analysis (PCA) on a selected portion of the image, wherein each resulting orthogonal basis set has “N” basis vectors; and with each iteration including:

puncturing of selected dimensional components of the orthogonal basis set resulting from each PCA of the selected portion of the image, without removing any of the associated “N” basis vectors thereof; and

wherein the iterative act further comprises:

transforming each of the “N” basis vectors in the prior orthogonal basis set into a corresponding Fourier series; and

puncturing higher frequency ones of the Fourier series coefficients associated with the “N” basis vectors;

removing selected basis vectors of a final one of the resulting orthogonal basis sets of the iteration act; thereby forming a truncated basis set for compression of the selected portion of the image; and

projecting the image onto the truncated basis set thereby obtaining coefficients of the selected portion of the image, relative to the truncated basis set.

11. The method for compressing an image of claim 10 , wherein the iterative act further comprises:

transforming each of the “N” basis vectors in the prior orthogonal basis set into a corresponding Fourier series;

quantizing a phase component of the coefficients of each of the “N” basis vector's associated Fourier series; and

puncturing higher frequency ones of the Fourier series coefficients associated with the “N” basis vectors.

12. A non-transient computer readable medium containing program instructions for causing a computer to perform the method of:

iteratively performing a Principal Component Analysis (PCA) on a selected portion of an image, wherein each resulting orthogonal basis set has “N” basis vectors; and with each iteration including:

puncturing of selected dimensional components of the orthogonal basis set resulting from each PCA of the selected portion of the image, without removing any of the associated “N” basis vectors thereof; and

conditionally iteratively performing the PCA based on a determination of a rank of an initial orthogonal basis set, whereby the initial orthogonal basis set exhibiting a low rank results in a negative determination with respect to the performance of the puncturing and subsequent PCA and vice versa;

removing selected basis vectors of a final one of the resulting orthogonal basis sets of the iteration act; thereby forming a truncated basis set for compression of the selected portion of the image; and

projecting the image onto the truncated basis set thereby obtaining coefficients of the selected portion of the image, relative to the truncated basis set.

13. A non-transient computer readable medium containing program instructions for causing a computer to perform the method of:

iteratively performing a Principal Component Analysis (PCA) on a selected portion of an image, wherein each resulting orthogonal basis set has “N” basis vectors; and with each iteration including:

transforming each of the “N” basis vectors in the prior orthogonal basis set into a corresponding Fourier series; and

puncturing of selected dimensional components of the orthogonal basis set resulting from each PCA of the selected portion of the image, without removing any of the associated “N” basis vectors thereof; including:

puncturing higher frequency ones of the Fourier series coefficients associated with the “N” basis vectors;

removing selected basis vectors of a final one of the resulting orthogonal basis sets of the iteration act; thereby forming a truncated basis set for compression of the selected portion of the image; and

projecting the image onto the truncated basis set thereby obtaining coefficients of the selected portion of the image, relative to the truncated basis set.

Assignments (9)