ESTIMATION OF POINT SPREAD FUNCTIONS FROM MOTION-BLURRED IMAGES

• US 20100231732A1
• Filed: 03/10/2010
• Published: 09/16/2010
• Est. Priority Date: 03/11/2009
• Status: Active Grant

• Current Assignee: Qualcomm, Inc.
• Original Assignee: Zoran Corporation (Qualcomm, Inc.)
• Inventors: TZUR, MEIR, BAXANSKY, ARTEMY
• Granted Patent: US 8,698,905 B2
• Time in Patent Office: Days
• US Class Current: 348/208.400
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/20192 Edge enhancement; Edge preservation
• CPC Class Codes: G06T 2207/20201 Motion blur correction
• CPC Class Codes: G06T 5/20 using local operators
• CPC Class Codes: G06T 5/73 Deblurring; Sharpening
• CPC Class Codes: H04N 23/68 for stable pick-up of the scene, e.g. compensating for camera body vibrations
• CPC Class Codes: H04N 23/6811 based on the image signal
• CPC Class Codes: H04N 23/683 performed by a processor, e.g. controlling the readout of an image memory
• CPC Class Codes: H04N 23/80 Camera processing pipelines; Components thereof

Abstract

Methods for estimating the point spread function (PSF) of a motion-blurred image are disclosed and claimed. In certain embodiments, the estimated PSF may be used to compensate for the blur caused by hand-shake without the use of an accelerometer or gyro. Edge spread functions may be extracted along different directions from straight edges in a blurred image and combined to find the PSF that best matches. In other embodiments, the blur response to edges of other forms may similarly be extracted, such as corners or circles, and combined to find the best matching PSF. The PSF may then be represented in a parametric form, where the parameters used are related to low-order polynomial coefficients of the angular velocity vx(t) and vy(t) as a function of time.

Claims

1. A method of de-blurring images by estimating point spread functions (PSFs), the method comprising the acts of:

receiving image data corresponding to a captured image;

identifying a plurality of shapes within the captured image;

estimating a plurality of shape spread functions corresponding to each of the plurality of shapes;

determining the best matching angular velocity parameters to said plurality of shape spread functions;

calculating the PSF from the determined angular velocity parameters; and

de-blurring said image data using the calculated PSF.

2. The method of claim 1, wherein the plurality of shapes are plurality of isolated edges having a corresponding plurality of directions and the plurality of shape spread functions are edge spread functions (ESFs) corresponding to each of the plurality of isolated edges.

3. The method of claim 2, wherein identifying the plurality of isolated edges comprises performing a template matching operation over a range of angles of inclination.

4. The method of claim 2, wherein each of the plurality of isolated edges has a width defined as a distance in pixels from 10% to 90% of an edge height in a direction of maximum change.

5. The method of claim 4, and wherein the plurality of isolated edges includes only isolated edges having widths close to a determined minimum.

6. The method of claim 2, wherein estimating the plurality of ESFs comprises aligning and averaging the cross-sections of each of the plurality of isolated edges in each of the plurality of directions.

7. The method of claim 2, further comprising:

comparing the widths of the plurality of ESFs among the plurality of directions;

excluding ESFs having widths that are inconsistent with the other ESFs.

8. The method of claim 1, wherein calculating the PSF comprises representing the PSF in parametric form using low-order polynomial coefficients of angular velocity.

9. A camera comprising:

an image capturing system; and

an image processing circuit coupled to the image capturing system, wherein the image processing circuit is to;

receive image data corresponding to a captured image;

identify a plurality of shapes within the captured image;

estimate a plurality of shape spread functions corresponding to each of the plurality of isolated edges;

determine the best matching angular velocity parameters to said plurality of shape spread functions;

calculate the PSF from the determined angular velocity parameters; and

de-blur said image data using the calculated PSF.

10. The camera of claim 9, wherein the plurality of shapes are plurality of isolated edges having a corresponding plurality of directions and the plurality of shape spread functions are edge spread functions (ESFs) corresponding to each of the plurality of isolated edges.

11. The camera of claim 10, wherein the image processing circuit is to identify the plurality of isolated edges by performing a template matching technique over a range of angles of inclination.

12. The camera of claim 10, wherein each of the plurality of isolated edges has a width defined as a distance in pixels from 10% to 90% of an edge height in a direction of maximum change.

13. The camera of claim 12, and wherein the plurality of isolated edges includes only isolated edges having widths close to a determined minimum.

14. The camera of claim 10, wherein the image processing circuit is further to estimate the plurality of ESFs by aligning and averaging the cross-sections of each of the plurality of isolated edges in each of the plurality of directions.

15. The camera of claim 10, wherein the image processing circuit is further to:

compare the widths of the plurality of ESFs among the plurality of directions; and

exclude ESFs having widths that are inconsistent with the other ESFs.

16. The camera of claim 9, wherein the image processing circuit the PSF is represented in parametric form using low-order polynomial coefficients of angular velocity.

17. A computer program product comprising:

a processor storage medium having computer executable program code embodied therein to de-blur images by estimating point spread functions (PSFs), the processor storage medium having;

computer executable program code to identify a plurality of shapes within image data corresponding to a captured image;

computer executable program code to estimate a plurality of shape spread functions corresponding to each of the plurality of shapes;

computer executable program code to determine the best matching angular velocity parameters to said plurality of shape spread functions;

computer executable program code to calculate the PSF from the determined angular velocity parameters; and

computer executable program code to de-blur said image data using the calculated PSF.

18. The computer program of claim 17, wherein the plurality of shapes are plurality of isolated edges having a corresponding plurality of directions and the plurality of shape spread functions are edge spread functions (ESFs) corresponding to each of the plurality of isolated edges.

19. The computer program product of claim 18, wherein the computer executable program code to identify the plurality of isolated edges comprises computer executable program code to perform a template matching operation over a range of angles of inclination.

20. The computer program product of claim 18, wherein each of the plurality of isolated edges has a width defined as a distance in pixels from 10% to 90% of an edge height in a direction of maximum change.

21. The computer program product of claim 20, and wherein the plurality of isolated edges includes only isolated edges having widths close to a determined minimum.

22. The computer program product of claim 18, wherein the computer executable program code to estimate the plurality of ESFs comprises computer executable program code to align and average the cross-sections of each of the plurality of isolated edges in each of the plurality of directions.

23. The computer program product of claim 18, further comprising computer executable program code to:

compare the widths of the plurality of ESFs among the plurality of directions;

exclude ESFs having widths that are inconsistent with the other ESFs.

24. The computer program product of claim 17, wherein the computer executable program code to calculate the PSF comprises computer executable program code to represent the PSF in parametric form using low-order polynomial coefficients of angular velocity.