Method and system for image compression using image symmetry
First Claim
Patent Images
1. A computer-implemented method for reducing an image representing a substantially symmetric subject, comprising the steps of:
- dividing the image into a first part and a second part, the second part being substantially symmetric to the first part;
determining pairs of symmetrically corresponding pixels, a member of the pair in the first part having an X pixel value and a member of the pair in the second part having a Y pixel value;
computing a weighted average value A of the pixel values of each one of the pairs;
computing a weighted variance value B of the pixel values of each one of the pairs;
representing each of the pairs with the weighted average and the weighted variance values;
generating reduced image data including at least a plurality of A values and corresponding B values the pairs;
reconstituting the image by computing the X pixel value and the Y pixel value from the weighted average value A and the weighted variance value B; and
compressing the reduced image data with an image compression algorithm, wherein the reconstituting step further includes the substep of removing visual artifacts from the reconstituted image, wherein the visual artifacts comprise a seam line between the first part and the second part, wherein the dividing, determining, computing, representing and generating steps are executed by a microprocessor.
4 Assignments
0 Petitions
Accused Products
Abstract
An image compression system and method is described, which makes use of the symmetry found in faces and heads to perform the compression. The image is divided along the line of symmetry, and pairs of corresponding pixels on the two divided sides are determined. A weighted average and a weighted variance of the pixel values of the pairs is computed, and is used to encode the image. A transform such as the Karhunen-Loeve transform is used to compute the weighted averages and variances.
-
Citations
17 Claims
-
1. A computer-implemented method for reducing an image representing a substantially symmetric subject, comprising the steps of:
-
dividing the image into a first part and a second part, the second part being substantially symmetric to the first part; determining pairs of symmetrically corresponding pixels, a member of the pair in the first part having an X pixel value and a member of the pair in the second part having a Y pixel value; computing a weighted average value A of the pixel values of each one of the pairs; computing a weighted variance value B of the pixel values of each one of the pairs; representing each of the pairs with the weighted average and the weighted variance values; generating reduced image data including at least a plurality of A values and corresponding B values the pairs; reconstituting the image by computing the X pixel value and the Y pixel value from the weighted average value A and the weighted variance value B; and compressing the reduced image data with an image compression algorithm, wherein the reconstituting step further includes the substep of removing visual artifacts from the reconstituted image, wherein the visual artifacts comprise a seam line between the first part and the second part, wherein the dividing, determining, computing, representing and generating steps are executed by a microprocessor. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)
-
-
17. A computer-implemented method for reducing an image representing a substantially symmetric subject, comprising the steps of:
-
dividing the image into a first part and a second part, the second part being substantially symmetric to the first part; determining pairs of symmetrically corresponding pixels, a member of the pair in the first part having an X pixel value and a member of the pair in the second part having a Y pixel value; computing a weighted average value A of the pixel values of each one of the pairs; computing a weighted variance value B of the pixel values of each one of the pairs; representing each of the pairs with the weighted average and the weighted variance values; generating reduced image data including at least a plurality of A values and corresponding B values the pairs; reconstituting the image by computing the X pixel value and the Y pixel value from the weighted average value A and the weighted variance value B; and compressing the reduced image data with an image compression algorithm, wherein the reconstituting step further includes the substep of removing visual artifacts from the reconstituted image, wherein the removing step includes the substeps of comparing the weighted variance value to a difference in weighted average values for adjacent pixels, wherein the dividing, determining, computing, representing and generating steps are executed by a microprocessor.
-
Specification