METHOD AND SYSTEM FOR REDUCING RINGING ARTIFACTS OF IMAGE DECONVOLUTION

• US 20110085743A1
• Filed: 09/29/2010
• Published: 04/14/2011
• Est. Priority Date: 10/13/2009
• Status: Active Grant

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Lei, Zhichun, Zimmermann, Klaus
• Granted Patent: US 8,588,544 B2
• Time in Patent Office: Days
• US Class Current: 382/264
• CPC Class Codes: G06T 2207/20192 Edge enhancement; Edge preservation
• CPC Class Codes: G06T 5/73 Deblurring; Sharpening

Abstract

The present invention relates to a method for reducing ringing artifacts of image deconvolution. A deconvolution is applied to the original blurred image and edges are detected within the deconvolved image. Then edge tapering on the original blurred image is performed based on the detected edges.

The present invention further relates to a system and a computer program product for reducing ringing artifacts of image deconvolution.

Claims

1. Method for reducing ringing artifacts of image deconvolution, comprising the steps of

applying a deconvolution to the original blurred image,

detecting edges within the deconvolved image and

performing edge tapering on the original blurred image based on the detected edges.

2. Method according to claim 1, further comprising the steps of

low pass filtering the deconvolved image,

downsampling the filtered image and

additionally using the downsampled filtered image for edge detection.

3. Method according to claim 2, comprising the step of repeating the low pass filtering and downsampling at least once to obtain downsampled filtered images with different resolution for edge detection.

4. Method according to claim 1, 2 or 3, wherein the step of detecting edges comprises

calculating a pixel difference between adjacent pixels and

determining that an edge is present if the pixel difference is exceeding a predefined threshold value.

5. Method according to claim 1, 2 or 3, comprising the step of applying a second deconvolution to the edge tapered image thereby obtaining a ringing-reduced image.

6. Method according to claim 5, comprising the step of repeating the step of edge detection by detecting edges within the ringing-reduced image and repeating the step of edge tapering on the original blurred image based on the edges detected within the ringing-reduced image.

7. Method according to claim 1, 2 or 3, wherein the deconvolution is iterative or non-iterative.

8. Method according to claim 1, 2 or 3, wherein the step of edge tapering comprises attenuating the pixels within the size or multiple size of the point spread function around the detected edges.

9. Method according to claim 8, wherein the attenuation is accomplished by multiplying the pixels with a positive fractional weighting factor w, wherein preferably 0<

w<

1.

14. System according to any of the preceding claims, wherein the deconvolution unit is adapted to apply a second deconvolution to the edge tapered image thereby obtaining a ringing-reduced image.

15. System according to claim 14, wherein the edge detector is adapted to repeat the step of edge detection by detecting edges within the ringing-reduced image and repeating the step of edge tapering on the original blurred image based on the edges detected within the ringing-reduced image.

10. System for reducing ringing artifacts of image deconvolution, comprising

a deconvolution unit that applies a deconvolution to the original blurred image,

an edge detector that detects edges within the deconvolved image and

an edge tapering unit that performs edge tapering on the original blurred image based on the detected edges.

11. System according to claim 10, further comprising a multi-resolution unit comprising

a low pass filter that low pass filters the deconvolved image, and

a donwnsampling unit that downsamples the filtered image,

wherein the edge detector is adapted to additionally use the downsampled filtered image for edge detection.

12. System according to claim 11, wherein the multi-resolution unit is adapted to repeat the low pass filtering and downsampling at least once to obtain downsampled filtered images with different resolution for edge detection.

13. System according to claim 10, 11 or 12, wherein edge detector is adapted to

calculate a pixel difference between adjacent pixels and

determine that an edge is present if the pixel difference is exceeding a predefined threshold value.

16. System according to claim 10, 11 or 12, wherein the deconvolution is iterative or non-iterative.

17. System according to claim 10, 11 or 12, wherein the edge tapering unit is adapted to attenuate the pixels within the size or multiple size of the point spread function around the detected edges.

18. System according to claim 17, wherein the edge tapering unit is adapted to accomplish attenuation by multiplying the pixels with a positive fractional weighting factor w, wherein preferably 0<

w<

1.

19. Device, preferably a still camera, a video camera or a display, comprising a system according to any of claim 10, 11 or 12.

20. System for reducing ringing artifacts of image deconvolution, comprising

means for applying a deconvolution to the original blurred image,

means for detecting edges within the deconvolved image and

means for performing edge tapering on the original blurred image based on the detected edges.

21. A computer program product stored on a computer readable non-transitory medium which causes a computer to perform the steps of

applying a deconvolution to the original blurred image,

detecting edges within the deconvolved image and

performing edge tapering on the original blurred image based on the detected edges.

22. Computer readable non-transitory storage medium comprising a computer program product according to claim 21.