SHARPNESS ESTIMATION IN DOCUMENT AND SCENE IMAGES

• US 20130121610A1
• Filed: 11/10/2011
• Published: 05/16/2013
• Est. Priority Date: 11/10/2011
• Status: Active Grant

• Current Assignee: Fujifilm Business Innovation Corp. (Fujifilm Holdings Corporation)
• Original Assignee: Fuji Xerox Company Limited (Xerox Holdings Corp.)
• Inventors: CHEN, Francine, KUMAR, Jayant
• Granted Patent: US 9,087,262 B2
• Time in Patent Office: Days
• US Class Current: 382/266
• CPC Class Codes: G06T 2207/30176 Document
• CPC Class Codes: G06T 7/0002 Inspection of images, e.g. flaw detection
• CPC Class Codes: G06V 30/10 Character recognition

Abstract

Systems and methods described herein are directed to estimating the sharpness of images or photos such as document pages or scenes. The systems and methods are based on calculating a difference in grayscale values between pixels at an edge of an image. The differences may be accomplished by taking a slope of grayscale values between two pixels at an edge window, and estimating the edge sharpness based on the slope, multiple slopes, or differences in the slopes.

Claims

1. An apparatus, comprising:

a camera that captures an image;

a smoothing unit that generates at least one smoothed image based on the captured image, over at least one of an x-axis direction and a y-axis direction;

an identifying unit that identifies a first window comprising a first plurality of pixels around an edge pixel of the at least one smoothed image, and a second window in the captured image comprising a second plurality of pixels around an edge pixel of the captured image, and corresponding to the first plurality of pixels in the first window; and

an estimating unit that estimates edge sharpness by determining differences in grayscale value between first pairs of pixels from the first window and second pairs of pixels from the second window over at least one of an x-axis direction and a y-axis direction of each of the first window and the second window.

2. The apparatus of claim 1, wherein the estimating unit estimates the edge sharpness by identifying a maximum difference of the differences in the grayscale value.

3. The apparatus of claim 1, wherein the estimating unit estimates the edge sharpness by summing the differences in the grayscale value between the first pairs of pixels and second pairs of pixels over the x-axis direction, and between the first pairs of pixels and the second pairs of pixels over the y-axis direction.

4. The apparatus of claim 1, wherein the smoothing unit generates the at least one smoothed image over the x-axis direction by utilizing a median filter over the x-axis for determining differences over the x-axis, and generates the at least one smoothed image over the y-axis direction by utilizing the median filter over the y-axis for determining differences over the y-axis.

5. The apparatus of claim 1, wherein the estimating unit estimates a number of sharp pixels from the differences in greyscale value and generates a score based on a ratio of a number of sharp pixels in a region of the captured image comprising the second window, and a number of pixels in the region belonging to an edge of the captured image.

6. The apparatus of claim 1, wherein the estimating unit normalizes the differences in the greyscale values by a function of gradient values of the second plurality of pixels in the second window.

7. A non-transitory computer readable medium storing instructions for execution by a processor, comprising:

receiving a captured image;

generating at least one smoothed image of the captured image over at least one of an x-axis direction and a y-axis direction;

identifying a first window comprising a first plurality of pixels around an edge pixel of the at least one smoothed image;

identifying a second window corresponding to the first window in the captured image comprising a second plurality of pixels around an edge pixel of the captured image, and corresponding to the first plurality of pixels in the first window; and

estimating edge sharpness by determining differences in grayscale value between first pairs of pixels from the first window and second pairs of pixels from the second window over at least one of an x-axis direction and a y-axis direction of each of the first window and the second window.

8. The non-transitory computer readable medium of claim 7, wherein the estimating the edge sharpness further comprises identifying a maximum difference of the differences in the grayscale value.

9. The non-transitory computer readable medium of claim 7, wherein the estimating the edge sharpness further comprises summing the differences in the grayscale value between the first pairs of pixels and second pairs of pixels over the x-axis direction, and between the first pairs of pixels and the second pairs of pixels over the y-axis direction.

10. The non-transitory computer readable medium of claim 7, further comprising generating the at least one smoothed image over the x-axis direction by utilizing a median filter over the x-axis for determining differences over the x-axis;

and

generating the at least one smoothed image over the y-axis direction by utilizing the median filter over the y-axis for determining differences over the y-axis.

11. The non-transitory computer readable medium of claim 7, further comprising estimating a number of sharp pixels from the differences in greyscale value and generating a score based on a ratio of a number of sharp pixels in a region of the captured image comprising the second window and a number of pixels in the region belonging to an edge of the captured image.

12. The non-transitory computer readable medium of claim 7, further comprising normalizing the differences in greyscale values by a function of gradient values of the second plurality of pixels in the second window.

13. A method, comprising:

utilizing a camera to capture an image;

generating at least one smoothed image of the captured image over at least one of an x-axis direction and a y-axis direction;

identifying a first window comprising a first plurality of pixels around an edge pixel of the at least one smoothed image;

identifying a second window corresponding to the first window in the captured image comprising a second plurality of pixels around an edge pixel of the captured image, and corresponding to the first plurality of pixels in the first window; and

estimating edge sharpness by determining differences in grayscale value between first pairs of pixels from the first window and second pairs of pixels from the second window over at least one of an x-axis direction and a y-axis direction of each of the first window and the second window.

14. The method of claim 13, wherein the estimating the edge sharpness further comprises identifying a maximum difference of the differences in grayscale value over the x-axis direction and over the y-axis direction.

15. The method of claim 13, wherein the estimating the edge sharpness further comprises summing differences in grayscale value between the first pairs of pixels and second pairs of pixels over the x-axis direction and between the first pairs of pixels and the second pairs of pixels over the y-axis direction.

16. The method of claim 13, further comprising generating the at least one smoothed image over the x-axis direction by utilizing a median filter over the x-axis for determining differences over the x-axis;

and

generating the at least one smoothed image over the y-axis direction by utilizing the median filter over the y-axis for determining differences over the y-axis.

17. The method of claim 13, further comprising estimating a number of sharp pixels from the differences in greyscale value and generating a score based on a ratio of a number of sharp pixels in a region of the captured image comprising the second window, and a number of pixels in the region belonging to an edge of the captured image.

18. The method of claim 13, further comprising normalizing the differences in the greyscale values by a function of gradient values of the second plurality of pixels in the second window.