Method and apparatus for deblurring non-uniform motion blur using multi-frame including blurred image and noise image

1. A method of deblurring a non-uniform motion blur, the method comprising:

receiving a multi-frame comprising a blurred image including a non-uniform motion blur resulting from motion other than a translational motion, and a noise image in which a blur is absent;

performing image brightness registration with respect to the blurred image and the noise image using histogram matching for correcting a difference in exposure times;

estimating non-uniform motion blur information of a blurred image generated by performing the image brightness registration; and

obtaining a latent image by deblurring the non-uniform motion blur in the multi-frame, based on the estimated non-uniform motion blur information.

2. The method of claim 1 , wherein the performing of the image brightness registration comprises:

removing noise from the noise image; and

performing image brightness registration with respect to the blurred image and the noise image from which the noise has been removed.

3. The method of claim 2 , further comprising:

performing the estimating of the non-uniform motion blur information and the obtaining of the latent image, iteratively.

4. The method of claim 3 , wherein the performing of the estimating iteratively comprises performing the estimating of the non-uniform motion blur information and the obtaining of the latent image, iteratively, while changing a resolution of the blurred image generated by performing the image brightness registration.

5. The method of claim 3 , further comprising:

estimating the non-uniform motion blur information of the noise image from which the noise has been removed during a first iteration.

6. The method of claim 3 , further comprising:

updating the non-uniform motion blur information during each of a second iteration and subsequent iterations using a latent image obtained during a previous iteration.

7. The method of claim 3 , wherein the obtaining of the latent image comprises obtaining a final latent image that is restored from the multi-frame by applying the Richardson-Lucy (RL) deconvolution algorithm with respect to final non-uniform motion blur information obtained by the performing of the estimating iteratively, the blurred image including the non-uniform motion blur, and the noise image from which the noise has been removed.

8. The method of claim 1 , wherein the estimating of the non-uniform motion blur information comprises:

estimating a homography for the blurred image on which the image brightness registration has been performed; and

computing a weight for the homography, using the estimated homography.

9. The method of claim 8 , wherein the estimating of the homography and the computing of the weight are performed a predetermined number of times, iteratively.

10. The method of claim 1 , wherein the obtaining of the latent image comprises deblurring the non-uniform motion blur using a residual blurred image and a residual latent image based on the RL deconvolution algorithm.

11. A non-transitory computer-readable medium comprising a program for instructing a computer to perform the method of claim 1 .

12. The method of claim 1 , wherein the latent image has less motion blur than the blurred image and the latent image has less noise than the noise image.

13. An apparatus for deblurring a non-uniform motion blur, the apparatus comprising:

a processor to control one or more processor-executable units;

a receiving unit to receive a multi-frame comprising a blurred image including a non-uniform motion blur resulting from motion other than a translational motion, and a noise image in which a blur is absent;

a performing unit to perform image brightness registration with respect to the blurred image and the noise image using histogram matching for correcting a difference in exposure times;

an estimating unit to estimate non-uniform motion blur information of a blurred image generated by performing the image brightness registration; and

an obtaining unit to obtain a latent image by deblurring the non-uniform motion blur in the multi-frame, based on the estimated non-uniform motion blur information.

14. The apparatus of claim 13 , wherein the performing unit comprises:

a noise removing device to remove noise from the noise image; and

a performing device to perform image brightness registration with respect to the blurred image and the noise image from which the noise has been removed.

15. The apparatus of claim 14 , further comprising:

a control unit to perform the estimating of the non-uniform motion blur information and the obtaining of the latent image, iteratively.

16. The apparatus of claim 15 , wherein the control unit performs the estimating of the non-uniform motion blur information and the obtaining of the latent image, iteratively, while changing a resolution of the blurred image generated by performing the image brightness registration.

17. The apparatus of claim 15 , wherein the control unit estimates non-uniform motion blur information of the noise image from which the noise has been removed, during a first iteration.

18. The apparatus of claim 15 , wherein the control unit updates the non-uniform motion blur information during each of a second iteration and subsequent iterations using a latent image obtained during a previous iteration.

19. The apparatus of claim 15 , wherein the obtaining unit obtains a final latent image that is restored from the multi-frame by applying the Richardson-Lucy (RL) deconvolution algorithm with respect to final non-uniform motion blur information obtained by the performing of the estimating iteratively, the blurred image including the non-uniform motion blur, and the noise image from which the noise has been removed.

20. The apparatus of claim 13 , wherein the estimating unit comprises:

an estimating device to estimate a homography for the blurred image on which the image brightness registration has been performed; and

a computing device to compute a weight for the homography, using the estimated homography.

21. The apparatus of claim 20 , wherein the estimating of the homography by the estimating device and the computing of the weight by the computing device are performed a predetermined number of times, iteratively.

22. The apparatus of claim 13 , wherein the latent image has less motion blur than the blurred image and the latent image has less noise than the noise image.

23. A method of deblurring a non-uniform motion blur, the method comprising:

receiving a multi-frame comprising a blurred image including a non-uniform motion blur resulting from motion other than a translational motion, and a noise image;

performing image brightness registration with respect to the blurred image and the noise image using histogram matching to correct a difference in exposure times;

estimating non-uniform motion blur information of a blurred image generated by performing the image brightness registration by estimating a homography for the blurred image on which the image brightness registration has been performed and computing a weight for the homography using the estimated homography; and

obtaining a latent image having less motion blur than the blurred image by deblurring the non-uniform motion blur in the multi-frame based on the estimated non-uniform motion blur information including the estimated homography and the computed weight.

24. The method of claim 23 , wherein the blurred image is obtained using a relatively long exposure time and the noise image is obtained using a relatively short exposure time.

25. The method of claim 24 , wherein the latent image has less noise than the noise image.

26. The method of claim 23 , wherein the estimating of the homography for the blurred image and the computing of the weight for the homography are iteratively performed a predetermined number of times.

27. The method of claim 23 , wherein the estimating of the homography for the blurred image and the computing of the weight for the homography are iteratively performed until a change in error value of a homography set is less than a predetermined threshold.