Variable blocking artifact size and offset detection

1. A system, comprising:

a computer;

a memory in the computer;

a content stored in the memory of the computer, the content having pixels arranged in two or more dimensions;

an edge difference calculator to calculate edge differences of the content in a first dimension of the two or more dimensions, the edge difference calculator being structured to calculate absolute differences within multiple sets of adjacent pixels in the content, the adjacent pixels within each of the multiple sets being adjacent in a dimension orthogonal to the first dimension and each pixel in each of the multiple sets of adjacent pixels being adjacent to a pixel in another set of the multiple sets, sum the absolute differences of the multiple sets of adjacent pixels to produce a sum, and scale the sum of the absolute differences by a number of pixels in the first dimension to produce the edge differences;

an edge difference filter to filter said edge differences to produce filtered edge differences; and

a block attribute calculator to determine a block attribute from said filtered edge differences.

2. A system according to claim 1 , wherein the edge difference filter is structured to calculate rolling standard deviations of the edge differences and filter out a first edge difference for which a corresponding first rolling standard deviation is less than a predetermined threshold greater than a second rolling standard deviation corresponding to a second edge difference immediately preceding said first edge difference.

3. A system according to claim 1 , wherein the block attribute calculator includes a block size calculator to determine a block size from said filtered edge differences.

4. A system according to claim 3 , wherein the block size calculator is structured to determine said block size as a most common difference between adjacent filtered edge differences.

5. A system according to claim 3 , wherein the block attribute calculator further includes a block offset calculator to determine a block offset from said filtered edge differences.

6. A system according to claim 5 , wherein the block offset calculator is structured to determine said block offset as a number of pixels between a filtered edge difference and an edge of the content.

7. A method, comprising:

calculating edge differences of a content of pixels arranged in two or more dimensions, the calculating edge differences including:

calculating absolute differences within multiple sets of adjacent pixels in the content in a first dimension of the two or more dimensions, each set of the multiple sets of adjacent pixels being orthogonal to the first dimension of the two or more dimensions and each pixel in each of the multiple sets of adjacent pixels being adjacent to a pixel in another set of the multiple sets;

summing the absolute differences of the multiple sets of adjacent pixels to produce a sum; and

scaling the sum by a number of pixels in the first dimension of the two or more dimensions to produce the edge differences;

filtering the edge differences to produce filtered edge differences; and

determining at least one block attribute from the filtered edge differences.

8. A method according to claim 7 , wherein:

the two or more dimensions includes N columns and M rows, where N and M are both greater than 1;

the calculating edge differences of a content includes:

calculating vertical edge differences of the M rows, in which each set of the multiple sets of adjacent pixels is orthogonal to a vertical dimension of the content when calculating the vertical edge differences, and

calculating horizontal edge differences of the N columns, in which each set of the multiple sets of adjacent pixels is orthogonal to a horizontal dimension of the content when calculating the horizontal edge differences;

the filtering the edge differences to produce filtered edge differences includes:

filtering the vertical edge differences to produce filtered vertical edge differences, and

filtering the horizontal edge differences to produce filtered horizontal edge differences; and

the determining at least one block attribute from the filtered edge differences includes:

determining at least a first block attribute from the filtered vertical edge differences, and

determining at least a second block attribute from the filtered horizontal edge differences.

9. A method according to claim 7 , wherein the filtering the edge differences to produce filtered edge differences includes:

calculating rolling standard deviations of the edge differences; and

filtering out a first edge difference for which a corresponding first rolling standard deviation is less than a predetermined threshold greater than a second rolling standard deviation corresponding to a second edge difference immediately preceding the first edge difference.

10. A method according to claim 7 , wherein the determining at least one block attribute from the filtered edge differences includes determining a block size from the filtered edge differences.

11. A method according to claim 10 , wherein the determining a block size from the filtered edge differences includes determining the block size as a most common difference between adjacent filtered edge differences.

12. A method according to claim 10 , wherein the determining at least one block attribute from the filtered edge differences further includes determining a block offset from the filtered edge differences.

13. A method according to claim 12 , wherein the determining a block offset from the filtered edge differences includes determining the block offset as a number of pixels between a filtered edge difference and an edge of the content.