DENOISING APPARATUS, SYSTEM AND METHOD

• US 20140003735A1
• Filed: 05/08/2013
• Published: 01/02/2014
• Est. Priority Date: 06/29/2012
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Voronov, German, Artyomov, Evgeny, Lee, Jong-Hwa
• Granted Patent: US 9,002,136 B2
• Time in Patent Office: Days
• US Class Current: 382/275
• CPC Class Codes: G06T 2207/20012 Locally adaptive
• CPC Class Codes: G06T 2207/20021 Dividing image into blocks, subimages or windows
• CPC Class Codes: G06T 5/70 Denoising; Smoothing

Abstract

A denoising apparatus comprising an image input unit which receives pixel data including color information of pixels included in a correction target image, a denoising unit which denoises the pixel data by a weight based averaging method, wherein the weight is set to a maximum value when a difference value between a correction target block and a comparison target block in the correction target image is zero, decreases linearly to zero as the difference value increases until it reaches a threshold value, and is set to zero when the difference value is greater than or equal to the threshold value, and an image output unit which outputs the pixel data processed by the denoising unit. The denoising unit assigns a corrected weight value to at least a guaranteed number of comparison target blocks for an impulse block, where an impulse block is a correction target block for which the number of non-zero weight valued comparison target blocks is less than a predetermined guaranteed number.

Claims

1. A denoising apparatus comprising:

an image input unit configured to receive pixel data including color information of pixels included in a correction target image;

a denoising unit configured to denoise the pixel data by a weighted-averaging method, wherein a weight is set to a maximum value when a difference value between a correction target block and a comparison target block in the correction target image is zero, decreases linearly to zero as the difference value increases until it reaches a threshold value, and is set to zero when the difference value is equal to or greater than the threshold value; and

an image output unit configured to output denoised pixel data,

wherein the denoising unit is configured to assign a corrected weight value to at least a guaranteed number of comparison target blocks for an impulse block, and the impulse block is a correction target block for which the number of non-zero weight valued comparison target blocks is less than a predetermined guaranteed number.

2. The denoising apparatus of claim 1, wherein if there is another non-zero weight valued impulse block among the comparison target blocks, the denoising unit is configured to assign the corrected weight value to some zero-weight valued general blocks among the comparison target blocks, and average the impulse block and the general blocks.

3. The denoising apparatus of claim 1, wherein the denoising unit is configured to sort the comparison target blocks for the impulse block based on their difference values, select up to the guaranteed number of comparison target blocks from those having smaller difference values as weight correction target blocks, and assign corrected weight values to the weight correction target blocks.

4. The denoising apparatus of claim 1, wherein the denoising unit is configured to classify the comparison target blocks for the impulse block based on their difference values, select at least the guaranteed number of comparison target blocks from those having smaller difference values based on the classification results as weight correction target blocks, and assign the corrected weight values to the weight correction target blocks.

5. The denoising apparatus of claim 4, wherein the denoising unit does not sort the comparison target blocks for the impulse block.

6. The denoising apparatus of claim 4, wherein the denoising unit is configured to assign the corrected weight value to zero-weight valued comparison target blocks of the weight correction target blocks.

7. The denoising apparatus of claim 1, wherein the denoising unit is configured to put the comparison target blocks for the impulse block into a plurality of bins corresponding to their difference values, wherein each bin is associated with a different range of difference values, and add the bins into a weight correction target bin one by one in the order of increasing difference value range of each bin until the number of comparison target blocks included in the weight correction target bin is greater than or equal to the guaranteed number.

8. The denoising apparatus of claim 7, wherein the plurality of bins include an internal bin having a difference value range from zero to the threshold value, and one or more external bins whose difference value ranges comprise values greater than the threshold value.

9. The denoising apparatus of claim 1, wherein the denoising apparatus is configured as an image signal processor (ISP).

10. A denoising method comprising:

inputting pixel data of a correction target image into a denoising apparatus; and

using the denoising apparatus to perform a weighted-average denoising process on the pixel data,

wherein said weighted-average denoising comprises;

setting a weight to a maximum value when a difference value between a correction target block and a comparison target block is zero, setting the weight to decrease linearly to zero as the difference value increases until it reaches a threshold value, and setting the weight to zero when the difference value is equal to or greater than the threshold value;

assigning a corrected weight value to at least a guaranteed number of comparison target blocks whose difference value is greater than the threshold value; and

averaging the correction target block and the comparison target blocks to which the corrected weight value has been assigned to reduce noise in the correction target block pixel data,

wherein a correction target block for which the number of non-zero weight valued comparison target blocks is less than a predetermined guaranteed number is an impulse block.

11. The denoising method of claim 10, wherein said assigning comprises:

sorting the comparison target blocks for the impulse block based on the difference value;

selecting up to the guaranteed number of comparison target blocks having smaller difference values as weight correction target blocks using the sorting result; and

assigning the corrected weight value to the weight correction target blocks.

12. The denoising method of claim 10, wherein said assigning a corrected weight value comprises:

providing a plurality of bins having different ranges of the difference value;

putting each comparison target block for the impulse block into one of the plurality of bins corresponding to its difference value; and

adding the plurality of bins one by one in order of increasing a difference value range of each bin into a weight correction target bin until the number of comparison target blocks included in the weight correction target bin is greater than or equal to the guaranteed number.

13. A method for denoising pixels in a digital image, said method comprising the steps of:

receiving pixel data of a correction target image, wherein said correction target image includes a correction target block of pixels to be denoised, and one or more comparison target blocks of pixels;

setting a weight to a maximum value when a difference value between the correction target block and a comparison target block is zero;

setting the weight to decrease linearly to zero as the difference value increases until it reaches a threshold value;

setting the weight to zero when the difference value is equal to or greater than the threshold value;

assigning a corrected weight value to at least a guaranteed number of comparison target blocks whose difference value is greater than the threshold value; and

averaging the correction target block and the comparison target blocks to which the corrected weight value has been assigned to reduce noise in the correction target block pixel data,

wherein a correction target block for which the number of non-zero weight valued comparison target blocks is less than a predetermined guaranteed number is an impulse block.

14. The denoising method of claim 13, further comprising, if there is another non-zero weight valued impulse block among the comparison target blocks, assigning the corrected weight value to some zero-weight valued general blocks among the comparison target blocks, and averaging the impulse block and the general blocks.

15. The denoising method of claim 13, wherein assigning a corrected weight value comprises:

sorting the comparison target blocks for the impulse block based on the difference value;

selecting up to the guaranteed number of comparison target blocks having smaller difference values as weight correction target blocks using the sorting result; and

assigning the corrected weight value to the weight correction target blocks.

16. The denoising method of claim 13, wherein assigning a corrected weight value comprises:

providing a plurality of bins having different ranges of the difference value;

putting each comparison target block for the impulse block into one of the plurality of bins corresponding to its difference value; and

adding the plurality of bins one by one in order of increasing difference value range of each bin into a weight correction target bin until the number of comparison target blocks included in the weight correction target bin is greater than or equal to the guaranteed number.

17. The denoising method of claim 16, wherein the plurality of bins include an internal bin having a difference value range from zero to the threshold value, and one or more external bins whose difference value ranges comprise values greater than the threshold value.

18. The denoising method of claim 17, further comprising:

receiving the range size of the external bins and the number of external bins.

19. The denoising method of claim 13, further comprising:

receiving at least one of the threshold value, the guaranteed number, and the corrected weight value as an input.

20. The denoising method of claim 19, further comprising:

storing the at least one of the threshold value, the guaranteed number, and the corrected weight value.