Weighted absolute difference based noise reduction method and apparatus

1. A method comprising:

determining absolute differences between pixels in an image;

summing the absolute differences into at least one sum;

applying at least one weight to the at least one sum to determine at least two weighted sums;

comparing the at least two weighted sums to a predetermined threshold; and

applying an image filter responsive to the comparing.

2. The method of claim 1 where determining includes taking the absolute differences of luminance or gray scale between the pixels in the image.

3. The method of claim 1 where determining includes taking the absolute differences on horizontal, vertical, diagonal, and anti-diagonal directions.

4. The method of claim 1 where comparing includes:

identifying a highest and lowest of the at least two weighted sums;

calculating a difference between the highest and the lowest of the at least two weighted sums; and

comparing the difference with the predetermined threshold.

5. The method of claim 4 comprising:

applying an averaging filter where the difference is less than the predetermined threshold; and

applying a median filter responsive to the lowest weighted sum where the difference is not less than the predetermined threshold.

6. A method comprising:

determining a plurality of differences between pixels in close proximity in an image;

summing the plurality of differences to produce a plurality of sums;

applying a weight factor from a plurality of weight factors to each of the sums to produce a plurality of weighted sums;

summing the plurality of weighted sums to produce a plurality of directional operators;

comparing the plurality of directional operators to a predetermined threshold; and

applying an image filter responsive to the comparing.

7. The method of claim 6 where determining comprises:

picking a sub image including a first target pixel and a plurality of pixels surrounding the first target pixel;

taking a difference between the first target pixel and each surrounding pixel in a predetermined direction to produce a first group of the plurality of differences;

summing the first group of the plurality of differences to produce a first group sum; and

applying a first of the plurality of weight factors to the first group sum to produce a first weighted group sum.

8. The method of claim 7 , comprising:

picking a second target pixel in the sub image and at least one other surrounding pixel;

taking a difference between the second target pixel and the at least one other surrounding pixel in the predetermined direction to produce a second group of the plurality of differences;

summing the second group of the plurality of differences to produce a second group sum;

applying a second weight factor to the second group sum to produce a second weighted group sum; and

summing the first weighted group sum and the second weighted group sum to produce a directional operator of the plurality of directional operators in the predetermined direction.

9. The method of claim 7 where picking the sub image includes picking a 3 by 3 or a 3 by 5 image.

10. The method of claim 8 where applying the weight factors includes applying the first weight factor as 21/16 and the second weight factor as 21/32.

11. The method of claim 8 where taking differences includes taking absolute differences.

12. A method comprising:

determining a directional difference between minimum and maximum directional operators in an image;

comparing the directional difference to a predetermined threshold;

applying an averaging filter where the directional difference is less than the predetermined threshold; and

applying a median filter responsive to the minimum directional operator where the directional difference is not less than the predetermined threshold.

13. A method for noise reduction comprising:

calculating a plurality of directional parameters in an image;

sorting the plurality of directional parameters into an ordered list (S 0 , S 1 , S 2 , S 3 ) where S 3 ≧S 2 ≧S 1 ≧S 0 ;

determining a line structure by analyzing the plurality of directional parameters; and

applying a filter responsive to the determining.

14. The method of claim 13 where determining includes:

calculating a difference between highest and lowest directional parameters; and

comparing the difference with a predetermined threshold.

15. The method of claim 13 where calculating includes:

calculating a vertical directional parameter responsive to a first and second weight;

calculating a horizontal directional parameter responsive to the first and second weight;

calculating a diagonal directional parameter responsive to a third and fourth weight; and

calculating an anti-diagonal directional parameter responsive to the third and fourth weight.

16. The method of claim 15 where calculating includes calculating responsive to the first weight as 21/16, the second weight as 21/32, and the third and fourth weights as 1.

17. The method of claim 13 where calculating includes calculating a horizontal, vertical, diagonal, and anti-diagonal directional parameters.

18. The method of claim 13

where determining comprises comparing a difference between a highest directional parameter and a lowest directional parameter to a predetermined threshold; and

where applying comprises using a smoothing filter if the difference is less than the predetermined threshold.

19. The method of claim 18 where applying includes using a horizontal median filter, a vertical median filter, a diagonal median filter, or an anti-diagonal median filter responsive to the lowest directional parameter.

20. The method of claim 18 where applying includes:

using a diagonal median filter responsive to the lowest and the highest directional parameters;

using the diagonal and a vertical median filter responsive to the lowest and the highest directional parameters;

using the diagonal and a horizontal median filter responsive to the lowest and the highest directional parameters;

using an anti-diagonal median filter responsive to the lowest and the highest directional parameters;

using the anti-diagonal and the vertical median filter responsive to the lowest and the highest directional parameters; and

using the anti-diagonal and the horizontal median filter responsive to the lowest and the highest directional parameters.

21. An apparatus comprising:

means for calculating a plurality of directional operators in an image;

means for sorting the directional operators into an ordered list (S 0 , S 1 , S 2 , S 3 ) where S 3 ≧S 2 ≧S 1 ≧S 0 ; and

means for applying a filter responsive to the plurality of directional operators.

22. The apparatus of claim 21 where the means for applying comprises:

means for taking a difference between a highest directional operator and a lowest directional operator; and

means for applying an averaging filter if the difference is less than a predetermined threshold.

23. The apparatus of claim 21 where the means for applying comprises:

means for applying an averaging filter responsive to a predetermined threshold; and

means for applying a median filter responsive to a lowest directional operator and the predetermined threshold.