Fast Approximations of Shift-Variant Blur
- 908 Downloads
Image deblurring is essential in high resolution imaging, e.g., astronomy, microscopy or computational photography. Shift-invariant blur is fully characterized by a single point-spread-function (PSF). Blurring is then modeled by a convolution, leading to efficient algorithms for blur simulation and removal that rely on fast Fourier transforms. However, in many different contexts, blur cannot be considered constant throughout the field-of-view, and thus necessitates to model variations of the PSF with the location. These models must achieve a trade-off between the accuracy that can be reached with their flexibility, and their computational efficiency. Several fast approximations of blur have been proposed in the literature. We give a unified presentation of these methods in the light of matrix decompositions of the blurring operator. We establish the connection between different computational tricks that can be found in the literature and the physical sense of corresponding approximations in terms of equivalent PSFs, physically-based approximations being preferable. We derive an improved approximation that preserves the same desirable low complexity as other fast algorithms while reaching a minimal approximation error. Comparison of theoretical properties and empirical performances of each blur approximation suggests that the proposed general model is preferable for approximation and inversion of a known shift-variant blur.
KeywordsBlur Deconvolution Inverse problems Image restoration PSF
This work has been supported by Project MiTiV funded by the French National Research Agency (ANR DEFI 09-EMER-008-01). It has been performed in part within the framework of the LABEX PRIMES (ANR-11-LABX-0063) of Université de Lyon, within the program “Investissements d’Avenir” (ANR-11-IDEX-0007) operated by the French National Research Agency (ANR). Rahul Mourya acknowledges a PhD Grant funded by the Région Rhône-Alpes.
- Almeida, M., & Almeida, L. (2009). Blind deblurring of foreground-background images. In 2009 16th IEEE International Conference on Image Processing (ICIP), IEEE, (pp. 1301–1304).Google Scholar
- Ben Hadj, S., Blanc-Feraud, L., Aubert, G., & Engler, G. (2013). Blind restoration of confocal microscopy images in presence of a depth-variant blur and poisson noise. In 2013 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, (pp. 915–919).Google Scholar
- Calvetti, D., Lewis, B., & Reichel, L. (2000). Restoration of images with spatially variant blur by the gmres method. In Proceedings of SPIE, (vol. 4116, p. 364).Google Scholar
- Campisi, P., & Egiazarian, K. (Eds.). (2007). Blind image deconvolution: theory and applications. Boca Raton: CRC Press.Google Scholar
- Chakrabarti, A., Zickler, T., & Freeman, W. (2010). Analyzing spatially-varying blur. In 2010 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, (pp. 2512–2519).Google Scholar
- Chuang, Y., Curless, B., Salesin, D., & Szeliski, R. (2001). A bayesian approach to digital matting. In Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2001, IEEE, (vol. 2, pp. II-264).Google Scholar
- Denis, L., Thiébaut, E., & Soulez, F. (2011). Fast model of space-variant blurring and its application to deconvolution in astronomy. In IEEE International Conference on Image Process (ICIP).Google Scholar
- Escande, P., & Weiss, P. (2014). Numerical computation of spatially varying blur operators a review of existing approaches with a new one. arXiv preprint arXiv:1404.1023.
- Frigo, M. (1999). A fast Fourier transform compiler. In ACM Sigplan Notices, (vol. 34, pp. 169–180).Google Scholar
- Goodman, J. (2008). Introduction to Fourier optics. New York: McGraw-hill.Google Scholar
- Greenspan, H., Belongie, S., Goodman, R., Perona, P., Rakshit, S., & Anderson, C.H. (1994). Overcomplete steerable pyramid filters and rotation invariance. In Proceedings CVPR’94, 1994 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 1994, IEEE, (pp. 222–228).Google Scholar
- Hirsch, M., Sra, S., Scholkopf, B., & Harmeling, S. (2010). Efficient filter flow for space-variant multiframe blind deconvolution. In IEEE Computer Vision Pattern Recognition, (pp 607–614).Google Scholar
- Levin, A. (2007). Blind motion deblurring using image statistics. Advances in Neural Information Processing Systems, 19, 841.Google Scholar
- Nocedal, J. (1980). Updating quasi-Newton matrices with limited storage. Mathematics of Computation, 35(151), 773–782. doi: 10.1090/S0025-5718-1980-0572855-7.
- Soulez, F., Denis, L., Tourneur, Y., & Thiébaut, E. (2012). Blind deconvolution of 3d data in wide field fluorescence microscopy. In 2012 9th IEEE International Symposium on Biomedical Imaging (ISBI), IEEE, (pp 1735–1738).Google Scholar
- Whyte, O., Sivic, J., Zisserman, A., & Ponce, J. (2010). Non-uniform deblurring for shaken images. In 2010 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, (pp 491–498).Google Scholar
- Zoran, D., & Weiss, Y. (2011). From learning models of natural image patches to whole image restoration. In 2011 IEEE International Conference on Computer Vision (ICCV), IEEE, (pp. 479–486).Google Scholar