Methods and devices for defective pixel detection

• US 20070091187A1
• Filed: 10/26/2005
• Published: 04/26/2007
• Est. Priority Date: 10/26/2005
• Status: Active Grant

• Current Assignee: NVIDIA Corporation
• Original Assignee: NVIDIA Corporation
• Inventors: Lin, Shang-Hung
• Granted Patent: US 8,571,346 B2
• Time in Patent Office: Days
• US Class Current: 348/246
• CPC Class Codes: H04N 25/683 by defect estimation performed on the scene signal, e.g. real time or on the fly detection

Abstract

In a method of image signal processing, defective pixels are determined on-the-fly in a digital image representation based on a comparison of a pixel under evaluation with its surrounding pixels, with reference to a known resolving capability of a lens-sensor arrangement that captured the digital image representation. In response to the determination of defective pixels, the defective pixels are corrected.

Claims

1. A method of image signal processing comprising:

determining defective pixels on-the-fly in a digital image representation based on a comparison of a pixel under evaluation with its surrounding pixels, with reference to a known resolving capability of a lens-sensor arrangement that captured said digital image representation; and

in response to said determination correcting said defective pixels.

2. The method of image signal processing in claim 1 further comprising receiving said digital image representation in the form of a Bayer pattern array.

3. The method of image signal processing of claim 1 further comprising obtaining parameters describing the resolving capabilities of said lens-sensor arrangement.

4. The method of image signal processing in claim 1 further comprising:

calculating an average value of same colored pixels surrounding said pixel under evaluation in a portion of said digital image representation; and

determining if said pixel under evaluation is in a flat image region or an edge image region.

5. The method of image signal processing of claim 4 wherein said average value is calculated utilizing a modulation transfer function that is tailored for dark regions of images.

6. The method of image signal processing of claim 1 wherein said correcting said defective pixels comprises replacing each said defective pixel with a replacement pixel comprising an average value of same colored pixels surrounding said defective pixel in a portion of said digital image representation.

7. The method of image signal processing of claim 1 wherein said comparison comprises a comparison to a modified modulation transfer function unique to said lens-sensor arrangement.

8. The method of image signal processing of claim 7 wherein said comparison comprises comparing said pixel under evaluation to a modulation transfer function tailored for pixels in flat regions of images if said pixel under evaluation is in said flat region.

9. The method of image signal processing of claim 7 wherein said comparison comprises comparing said pixel under evaluation to a modulation transfer function tailored for pixels in edge regions of images if said pixel under evaluation is in said edge region.

10. An image signal processor device comprising:

a surrounding pixel interpolator for calculating an average value of same colored pixels surrounding a pixel under evaluation in a portion of a digital image representation;

a pixel evaluator coupled to said surrounding pixel interpolator, said pixel evaluator for determining if said pixel under evaluation is defective based on a comparison of said pixel under evaluation with its surrounding pixels, with reference to a known resolving capability of a lens-sensor arrangement that captured said digital image representation; and

a pixel replacer coupled to said pixel evaluator, wherein said pixel replacer is configured for replacing each defective pixel with a replacement pixel comprising said average value calculated by said surrounding pixel interpolator.

11. The image signal processor device of claim 10 further comprising a pixel location determiner coupled to said pixel evaluator, said pixel location determiner for determining if said pixel under evaluation is in a flat image region or an edge image region.

12. The image signal processor device of claim 10 further comprising an image data receiver coupled to said surrounding pixel interpolator, wherein said image data receiver is configured for receiving said digital image representation in the form of a Bayer pattern array.

13. The image signal processor device of claim 10 further comprising a memory element that comprises parameters that represent resolving capabilities of said lens-sensor arrangement.

14. The image signal processor device of claim 10 wherein said surrounding pixel interpolator is configured for calculating said average value utilizing a modulation transfer function that is tailored for dark regions of images.

15. The image signal processor device of claim 10 wherein said pixel evaluator is configured for determining if said pixel under evaluation is defective based on a comparison to a modified modulation transfer function unique to said lens-sensor arrangement.

16. The image signal processor device of claim 15 wherein said pixel evaluator is configured to utilize a modulation transfer function tailored for pixels in flat regions of images when said pixel under evaluation is in said flat image region.

17. The image signal processor device of claim 15 wherein said pixel evaluator is configured to utilize a modulation transfer function tailored for pixels in edge regions of images when said pixel under evaluation is in said edge image region.

18. A digital imaging device comprising:

an optical lens;

an image sensor coupled to said lens, said image sensor for capturing a digital image representation; and

an image signal processor coupled to said image sensor, said image signal processor comprising;

a surrounding pixel interpolator for calculating an average value of same colored pixels surrounding a pixel under evaluation in a portion of said digital image representation;

a pixel evaluator coupled to said surrounding pixel interpolator, said pixel evaluator for determining if said pixel under evaluation is defective based on a comparison of said pixel under evaluation with its surrounding pixels, with reference to a known resolving capability of a lens-sensor arrangement that captured said digital image representation; and

a pixel replacer coupled to said pixel evaluator, wherein said pixel replacer is configured for replacing each defective pixel with a replacement pixel comprising said average value calculated by said surrounding pixel interpolator.

19. The image signal processor device of claim 18 further comprising a pixel location determiner coupled to said pixel evaluator, said pixel location determiner for determining if said pixel under evaluation is in a flat image region or an edge image region.

20. The image signal processor device of claim 18 further comprising an image data receiver coupled to said surrounding pixel interpolator, wherein said image data receiver is configured for receiving said digital image representation in the form of a Bayer pattern array.