METHOD AND SYSTEM FOR NOISE MANAGEMENT FOR SPATIAL PROCESSING IN DIGITAL IMAGE/VIDEO CAPTURE SYSTEMS

• US 20090129695A1
• Filed: 09/23/2008
• Published: 05/21/2009
• Est. Priority Date: 11/15/2007
• Status: Active Grant

• Current Assignee: Marvell Asia Pte Limited (Marvell Technology Group Limited)
• Original Assignee: Marvell World Trade Limited (Marvell Technology Group Limited)
• Inventors: Hartin, Thomas, Aldrich, Bradley C., George, Adam
• Granted Patent: US 8,175,378 B2
• Time in Patent Office: Days
• US Class Current: 382/261
• CPC Class Codes: H04N 23/81 for suppressing or minimising disturbance in the image signal generation
• CPC Class Codes: H04N 23/843 Demosaicing, e.g. interpolating colour pixel values
• CPC Class Codes: H04N 25/134 based on three different wavelength filter elements

Abstract

A computer-implemented method for noise management in a digital image system measures noise levels of pixel data. The noise levels are adjusted with at least one of an intensity gain setting, a spatial gain setting, and a global gain setting to calculate noise adaptive thresholds for use during spatial processing of the pixel data.

Claims

1. A method, comprising:

measuring noise levels of pixel data; and

adjusting the noise levels with at least one of an intensity gain setting, a spatial gain setting, and a global gain setting to calculate noise adaptive thresholds for use during spatial processing of the pixel data.

2. The method of claim 1 further including outputting the noise adaptive thresholds.

3. The method of claim 2 further including outputting the noise adaptive thresholds to edge adaptive algorithms, wherein the edge adaptive algorithms comprise at least one of cross talk reduction, color synthesis, color management, and edge adaptive spatial filtering.

4. The method of claim 1 further including calculating the noise adaptive thresholds by combining the intensity gain setting, the spatial gain setting, and the global gain setting with dark pixel statistics.

5. The method of claim 4 further including reading the dark pixel statistics in one of real-time and from a calibration file.

6. The method of claim 4 further including combining the dark pixel statistics with the global gain setting and the intensity gain setting to create combined noise levels, and multiplying the combined noise levels by the spatial gain setting.

7. The method of claim 6 wherein the combined noise levels comprises a separate combined noise level calculated for each R, G, and B channel of the pixel data, and a separate noise adaptation threshold is calculated for each of the R, G, and B channels of the pixel data.

8. The method of claim 1 further including measuring the noise levels by calculating dark pixel statistics associated with dark pixel values of an image sensor, wherein the dark pixel statistics include at least one of a mean deviation, a standard deviation, and/or a variance of the dark pixel values.

9. The method of claim 8 further including calculating the noise adaptive thresholds as a product of the dark pixel statistics, the global gain setting, the intensity gain setting, and the spatial gain setting.

10. The method of claim 9 further including calculating the noise adaptive thresholds prior to initiating an image capture.

11. The method of claim 8 further including calculating the noise adaptive thresholds by,

calculating combined noise levels as a product of pixel statistics, the global gain setting, and the intensity gain setting prior to image capture;

and

calculating the noise adaptive thresholds as a product of the combined noise levels and the spatial gain setting.

12. The method of claim 11 further including calculating the noise adaptive thresholds after image capture.

13. The method of claim 11 wherein the combined noise levels comprise a combined noise level for each of R G and B channel that are calculated as a product of a global automatic exposure (AE) gain, analog and digital automatic white balance (AWB) gains for each of the R G and B channels, the intensity gain, and the standard deviation of the dark pixel values for each of the R G and B channels.

14. The method of claim 1 wherein the intensity gain setting is used during linearization correction, the spatial gain setting is used during vignetting correction, and the global gain setting is used by automatic white balance (AWB) and automatic exposure (AE) processes.

15. The method of claim 1 further including adjusting the noise levels and the noise adaptive thresholds based on a gain associated with nonlinearity of an image sensor.

16. A noise management system, comprising:

a statistics collection unit that measures noise levels of pixel data; and

a noise management unit that adjusts the noise levels with at least one of an intensity gain setting, a spatial gain setting, and a global gain setting to calculate noise adaptive thresholds for use during spatial processing of the pixel data.

17. The noise management system of claim 16 wherein the noise management unit outputs the noise adaptive thresholds.

18. The noise management system of claim 17 wherein the noise management unit outputs the noise adaptive thresholds to edge adaptive algorithms, wherein the edge adaptive algorithms comprise at least one of cross talk reduction, color synthesis, color management, and edge adaptive spatial filtering.

19. The noise management system of claim 16 wherein the noise management unit calculates the noise adaptive thresholds by combining the intensity gain setting, the spatial gain setting, and the global gain setting with dark pixel statistics.

20. The noise management system of claim 19 wherein the dark pixel statistics are read in one of real-time and from a calibration file.

21. The noise management system of claim 19 wherein the noise management unit combines the dark pixel statistics with the global gain setting and the intensity gain setting to create combined noise levels, and multiplies the combined noise levels by the spatial gain setting.

22. The noise management system of claim 21 wherein the combined noise levels comprises a separate combined noise level calculated for each R, G, and B channel of the pixel data, and a separate noise adaptation threshold is calculated for each of the R, G, and B channels of the pixel data.

23. The noise management system of claim 16 wherein the statistics collection unit measures the noise levels by calculating dark pixel statistics associated with dark pixel values of an image sensor, wherein the dark pixel statistics include at least one of a mean deviation, a standard deviation, and/or a variance of the dark pixel values.

24. The noise management system of claim 23 wherein the noise management unit calculates the noise adaptive thresholds as a product of the dark pixel statistics, the global gain setting, the intensity gain setting, and the spatial gain setting.

25. The noise management system of claim 24 wherein the noise management unit calculates the noise adaptive thresholds prior to initiating an image capture.

26. The noise management system of claim 23 wherein the noise management unit calculates the noise adaptive thresholds by,

calculating combined noise levels as a product of pixel statistics, the global gain setting, and the intensity gain setting prior to image capture;

and

calculating the noise adaptive thresholds as a product of the combined noise levels and the spatial gain setting.

27. The noise management system of claim 26 wherein the noise management unit calculates the noise adaptive thresholds after image capture.

28. The noise management system of claim 26 wherein the combined noise levels comprise a combined noise level for each of R G and B channel that are calculated as a product of a global automatic exposure (AE) gain, analog and digital automatic white balance (AWB) gains for each of the R G and B channels, the intensity gain, and the standard deviation of the dark pixel values for each of the R G and B channels.

29. The noise management system of claim 16 wherein the intensity gain setting is used during linearization correction, the spatial gain setting is used during vignetting correction, and the global gain setting is used by automatic white balance (AWB) and automatic exposure (AE) processes.

30. The noise management system of claim 16 wherein the noise management unit further adjusts the noise levels and the noise adaptive thresholds based on a gain associated with nonlinearity of an image sensor.