AUTOMATICALLY DETERMINING THE SIZE OF A BLUR KERNEL

• US 20160188998A1
• Filed: 02/16/2016
• Published: 06/30/2016
• Est. Priority Date: 10/23/2013
• Status: Active Grant

• Current Assignee: Adobe Inc.
• Original Assignee: Adobe Systems Incorporated (Adobe Inc.)
• Inventors: Cho, Sunghyun, Wang, Jue, Chien, Jen-Chan, Feng, Dong
• Granted Patent: US 9,576,220 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 18/23 Clustering techniques
• CPC Class Codes: G06T 2207/20024 Filtering details
• CPC Class Codes: G06T 2207/20048 Transform domain processing
• CPC Class Codes: G06T 2207/20172 Image enhancement details
• CPC Class Codes: G06T 2207/20192 Edge enhancement; Edge preservation
• CPC Class Codes: G06T 5/00 Image enhancement or restoration
• CPC Class Codes: G06T 5/73 Deblurring; Sharpening
• CPC Class Codes: G06T 7/60 Analysis of geometric attributes

Abstract

A computer-implemented method and apparatus are described for deblurring an image. The method may include accessing the image that has at least one blurred region and, automatically, without user input, determining a first value for a first size for a blur kernel for the at least one blurred region. Thereafter, automatically, without user input, a second value for a second size for the blur kernel is determined for the at least one blurred region. A suggested size for the blur kernel is then determined based on the first value and the second value.

Claims

1. A computer-implemented method, comprising:

generating a derivative image from an image having a blurred region;

autocorrelating the derivative image and the image to obtain an autocorrelation map;

determining a size of a blur kernel for the image based on the autocorrelation map; and

deblurring the image by deconvolving the blurred region with the blur kernel.

2. The method of claim 1, further comprising:

removing one or more nonzero elements less than a threshold value from the autocorrelation map.

3. The method of claim 2, further comprising:

identifying a connected component of nonzero elements in the autocorrelation map.

4. The method of claim 3, wherein determining the size of the blur kernel is based on the connected component of nonzero elements.

5. The method of claim 3, further comprising:

measuring one or more dimensions of the connected component of nonzero elements in the autocorrelation map.

6. The method of claim 3, further comprising:

measuring a maximum width or a maximum height of the connected component of nonzero elements in the autocorrelation map.

7. The method of claim 6, wherein determining the size of the blur kernel is based on the maximum width or the maximum height of the connected component.

8. The method of claim 1, wherein generating the derivative image comprises calculating a first order derivative or a second order derivative of the image.

9. The method of claim 8, further comprising:

removing pixels in flat regions of the image from being considered in the autocorrelation map based on the first order derivative or the second order derivative of the image.

10. The method of claim 1, wherein autocorrelating comprises computing the autocorrelation map with a Fourier transform.

11. The method of claim 1, wherein the size of the blur kernel is a first size, the method further comprising:

receiving input identifying a user-selected size for the blur kernel; and

determining a suggested size of the blur kernel based on the first size and the user-selected size.

12. An image deblurring system, comprising:

a processor;

an image module, coupled with the processor, the image module operated on the processor, to access an image having at least one blurred region; and

a blur kernel sizing module, coupled with the processor and the image module, to generate a derivative image from the image, autocorrelate the derivative image and the image to obtain an autocorrelation map, and determine a size of a blur kernel for the image based on the autocorrelation map.

13. The system of claim 12, wherein the blur kernel sizing module further comprises an autocorrelation-based size estimator to apply a threshold filter to the autocorrelation map to remove one or more nonzero elements less than a threshold value.

14. The system of claim 13, wherein the autocorrelation-based size estimator is further to measure a cluster of nonzero elements within the autocorrelation map.

15. The system of claim 14, wherein the autocorrelation-based size estimator is further to identify a connected component of nonzero elements in the autocorrelation map based on the cluster of nonzero elements.

16. The system of claim 15, wherein the autocorrelation-based size estimator is further to measure a maximum width or a maximum height of the connected component of nonzero elements in the autocorrelation map.

17. The system of claim 16, wherein the blur kernel sizing module is further to determine the size of the blur kernel based on the maximum width or the maximum height of the connected component.

18. The system of claim 12, wherein the blur kernel sizing module is further to remove pixels in flat regions of the image from being considered in the autocorrelation map based on a first order derivative or a second order derivative of the image.

19. The system of claim 12, wherein the blur kernel sizing module is further to autocorrelate the derivative image and the image based on a Fourier transform.

20. The system of claim 12, wherein the blur kernel sizing module is further to receive an input identifying a user-selected size for the blur kernel and determine the size of the blur kernel based on the autocorrelation map and the user-selected size for the blur kernel.