Noise reduction method and unit for an image frame

1. A method for noise reduction in a current image frame, wherein the current image frame is a part of a sequence of image frames, comprising:

comparing a current pixel value of a pixel within the current image frame with a corresponding pixel value of the pixel in at least one adjacent image frame;

determining a piled-up value of the current pixel value and the corresponding pixel value if a pixel difference between the current pixel value and the corresponding pixel value is below a pixel threshold, wherein the pixel threshold depends on the corresponding pixel value and/or the current pixel value;

dividing the current image frame and the at least one adjacent image frame into a plurality of blocks;

determining a current block piled-up value based on current pixel values of a respective block, which comprises the pixel;

determining a corresponding block piled-up value based on corresponding pixel values the respective block;

comparing the current block piled-up value with the corresponding block piled-up value; and

determining the piled-up value if a block difference between the current block piled-up value and the former block piled-up value is below a predetermined block threshold.

2. A method according to claim 1 , further comprising:

determining a block weight factor, depending on the block difference;

wherein in the determining a piled-up value the block weight factor is used to decrease an influence of the corresponding pixel value on the piled-up value.

3. A method according to claim 1 , wherein the block piled-up value is a sum of absolute differences of the pixel values of the pixels of the block.

4. A method for noise reduction in a current image frame, wherein the current image frame is a part of a sequence of image frames, comprising:

comparing a current pixel value of a pixel within the current image frame with a corresponding pixel value of the pixel in at least one adjacent image frame; and

determining a piled-up value of the current pixel value and the corresponding pixel value if a pixel difference between the current pixel value and the corresponding pixel value is below a pixel threshold, wherein the pixel threshold depends on the corresponding pixel value and/or the current pixel value;

wherein a plurality of pixel values for each pixel is provided, the plurality of pixel values describing a color value at the pixel in a color space model, wherein piled-up values for each pixel value of the plurality of pixel values are determined, if a color difference between each pixel value of the plurality of pixel values for a pixel is below a respective threshold; the method further comprising:

determining a color weight factor, depending on the color difference;

wherein in the determining a piled-up value the color weight factor is used to decrease an influence of the corresponding pixel value on the piled-up value.

5. A method according to claim 4 , wherein a luminance value, a first chrominance value, and a second chrominance value of a color space are used as respective pixel values of the plurality of pixel values.

6. A method for noise reduction in a current image frame, wherein the current image frame is a part of a sequence of image frames, comprising:

comparing a current pixel value of a pixel within the current image frame with a corresponding pixel value of the pixel in at least one adjacent image frame; and

determining a piled-up value of the current pixel value and the corresponding pixel value if a pixel difference between the current pixel value and the corresponding pixel value is below a pixel threshold, wherein the pixel threshold depends on the corresponding pixel value and/or the current pixel value;

wherein a plurality of pixel values for each pixel is provided, the plurality of pixel values describing a color value at the pixel in a color space model, wherein piled-up values for each pixel value of the plurality of pixel values are determined, if a color difference between each pixel value of the plurality of pixel values for a pixel is below a respective threshold; the method further comprising:

determining an accumulated weight factor based on the difference of pixel values, based on the difference of block piled-up values and based on the difference of color values;

wherein in the determining a piled-up value the accumulated weight factor is used to decrease an influence of the corresponding pixel value on the piled-up value.

7. A method according to claim 4 , further comprising:

up-sampling the color weight factor to a resolution of the pixel weight factor, if a resolution of the color weight value is lower than the resolution of the pixel weight value.

8. A method for noise reduction in a current image frame, wherein the current image frame is a part of a sequence of image frames, comprising:

comparing a current pixel value of a pixel within the current image frame with a corresponding pixel value of the pixel in at least one adjacent image frame;

determining a piled-up value of the current pixel value and the corresponding pixel value if a pixel difference between the current pixel value and the corresponding pixel value is below a pixel threshold, wherein the pixel threshold depends on the corresponding pixel value and/or the current pixel value;

storing a pre-sequence of image frames of a not-moving object;

generating a noise-reduced image by averaging pixel values of one pixel of the pre-sequence of image frames;

determining a variance of pixel values by comparing pixel values of the pixels for the pre-sequence of image frames; and

determining the pixel threshold from the variance.

9. A method for noise reduction in a current image frame, wherein the current image frame is a part of a sequence of image frames, comprising:

comparing a current pixel value of a pixel within the current image frame with a corresponding pixel value of the pixel in at least one adjacent image frame;

determining a piled-up value of the current pixel value and the corresponding pixel value if a pixel difference between the current pixel value and the corresponding pixel value is below a pixel threshold, wherein the pixel threshold depends on the corresponding pixel value and/or the current pixel value;

determining a difference value of piled-up values of the pixels of the image frame; and

performing a spatial noise reduction between piled-up values of adjacent pixels of the image frame if the difference value is below a difference threshold.

10. A method according to claim 9 , wherein the difference value is a variance value.

11. A method according to claim 9 , wherein an amount of the spatial noise reduction is based on the difference value.

12. A method according to claim 6 , further comprising:

determining textured areas and homogenous areas in the current image frame; and

determining the accumulated weight factor based on the pixel difference and the block difference for textured areas.

13. A method according to claim 6 , further comprising:

determining textured areas and homogenous areas in the current image frame; and

determining the accumulated weight factor for homogenous areas based on the pixel difference, the block difference, and the color difference.

14. A method for noise reduction in a current image frame, wherein the current image frame is a part of a sequence of image frames, comprising:

comparing a current pixel value of a pixel within the current image frame with a corresponding pixel value of the pixel in at least one adjacent image frame;

determining a piled-up value of the current pixel value and the corresponding pixel value if a pixel difference between the current pixel value and the corresponding pixel value is below a pixel threshold, wherein the pixel threshold depends on the corresponding pixel value and/or the current pixel value;

storing weight factors for a plurality of image frames of the sequence;

wherein the determining the piled-up value is based on the weight factors and the pixel values of the plurality of image frames.

15. A method for noise reduction in a current image frame, wherein the current image frame is a part of a sequence of image frames, comprising:

comparing a current pixel value of a pixel within the current image frame with a corresponding pixel value of the pixel in at least one adjacent image frame;

determining a piled-up value of the current pixel value and the corresponding pixel value if a pixel difference between the current pixel value and the corresponding pixel value is below a pixel threshold, wherein the pixel threshold depends on the corresponding pixel value and/or the current pixel value; and

storing the piled-up value of preceding image frame;

wherein the determining the piled-up value for a current image frame is based on the piled-up value of the preceding image frame.

16. A noise reduction unit for reducing noise in a current image frame, comprising:

a pixel comparator configured to compare a current pixel value of a pixel within the current image frame with a corresponding pixel value of the pixel in at least one adjacent image frame;

a processor, configured to determine a piled-up value of the current pixel value and the corresponding pixel value if a pixel difference between the current pixel value and corresponding pixel value is below a pixel threshold, wherein the pixel threshold depends on the corresponding pixel value and/or the current pixel value; and

a block comparator configured to divide the current image frame and the at least one adjacent image frame into a plurality of blocks, and to determine a current block piled-up value from the current pixel values of a respective block of the current image frame, which comprises the pixel, and to determine a corresponding block piled-up value from the corresponding pixel values of the respective block of the adjacent image frame, which comprises the pixel, and to compare the current block piled-up value with the corresponding block piled-up value.

17. A noise reduction unit for reducing noise in a current image frame, comprising:

a pixel comparator configured to compare a current pixel value of a pixel within the current image frame with a corresponding pixel value of the pixel in at least one adjacent image frame; and

a processor, configured to determine a piled-up value of the current pixel value and the corresponding pixel value if a pixel difference between the current pixel value and corresponding pixel value is below a pixel threshold, wherein the pixel threshold depends on the corresponding pixel value and/or the current pixel value; and

a block comparator configured to receive a current block piled-up value for the current image frame, to receive a corresponding block piled-up value for the adjacent image frame and being further configured to compare the current block piled-up value with the corresponding block piled-up value.

18. A noise reduction unit according to claim 16 , wherein the block comparator is further configured to determine a block weight factor, depending on a block difference between the current block piled-up value and the corresponding block piled-up value; and the processor is further configured to determine the piled-up value based on the block weight factor to decrease the influence of the corresponding pixel value on the piled-up value.

19. A noise reduction unit for reducing noise in a current image frame, comprising:

a pixel comparator configured to compare a current pixel value of a pixel within the current image frame with a corresponding pixel value of the pixel in at least one adjacent image frame;

a processor, configured to determine a piled-up value of the current pixel value and the corresponding pixel value if a pixel difference between the current pixel value and corresponding pixel value is below a pixel threshold, wherein the pixel threshold depends on the corresponding pixel value and/or the current pixel value; and

a color comparator configured to compare a current color value of a pixel within the current image frame with a corresponding color value of the pixel in at least one adjacent image frame;

wherein the processor is further configured determine the piled-up value if a color difference between the current color value and the corresponding color value is below a color threshold.

20. A noise reduction unit according to claim 19 , wherein the color comparator is further configured to determine a color weight factor, depending on a color difference between the current color value and the corresponding color value; and the processor is further configured to determine the piled-up value based on the color weight factor to decrease the influence of the corresponding pixel value on the piled-up value.

21. A noise reduction unit for reducing noise in a current image frame, comprising:

a pixel comparator configured to compare a current pixel value of a pixel within the current image frame with a corresponding pixel value of the pixel in at least one adjacent image frame;

a processor, configured to determine a piled-up value of the current pixel value and the corresponding pixel value if a pixel difference between the current pixel value and corresponding pixel value is below a pixel threshold, wherein the pixel threshold depends on the corresponding pixel value and/or the current pixel value; and

a noise concealment mechanism configured to determine a difference value of piled-up values of the pixels of the image frame and to perform a spatial noise reduction between piled-up values of adjacent pixels of the image frame if the difference value is below a difference threshold.

22. An image capturing device, comprising:

an optical set-up, configured to generate a sequence of image frames; and

a noise reduction unit according to any one of claims 16 , 17 , 19 and 21 .

23. An image displaying device, comprising:

a display; and

a noise reduction unit according to any one of claims 16 , 17 , 19 , and 21 .

24. A computer readable medium product comprising computer executable program code, which when loaded into a processor is configured to perform a method according to any one of claims 1 , 4 , 6 , 8 , 9 , 14 and 15 .