How to Compare Noisy Patches? Patch Similarity Beyond Gaussian Noise
- 1.6k Downloads
Many tasks in computer vision require to match image parts. While higher-level methods consider image features such as edges or robust descriptors, low-level approaches (so-called image-based) compare groups of pixels (patches) and provide dense matching. Patch similarity is a key ingredient to many techniques for image registration, stereo-vision, change detection or denoising. Recent progress in natural image modeling also makes intensive use of patch comparison.
A fundamental difficulty when comparing two patches from “real” data is to decide whether the differences should be ascribed to noise or intrinsic dissimilarity. Gaussian noise assumption leads to the classical definition of patch similarity based on the squared differences of intensities. For the case where noise departs from the Gaussian distribution, several similarity criteria have been proposed in the literature of image processing, detection theory and machine learning.
By expressing patch (dis)similarity as a detection test under a given noise model, we introduce these criteria with a new one and discuss their properties. We then assess their performance for different tasks: patch discrimination, image denoising, stereo-matching and motion-tracking under gamma and Poisson noises. The proposed criterion based on the generalized likelihood ratio is shown to be both easy to derive and powerful in these diverse applications.
KeywordsPatch similarity Likelihood ratio Detection Matching
Unable to display preview. Download preview PDF.
- Baxter, J., & Bartlett, P. (1998). The canonical distortion measure in feature space and 1-nn classification. In Advances in neural information processing systems 10: proceedings of the 1997 conference (Vol. 10, p. 245). Cambridge: MIT Press. Google Scholar
- Boykov, Y., Veksler, O., & Zabih, R. (1998). Markov random fields with efficient approximations. In Computer vision and pattern recognition, 1998. Proceedings. 1998 IEEE computer society conference on (pp. 648–655). New York: IEEE. Google Scholar
- Boykov, Y., Veksler, O., & Zabih, R. (2001). Fast approximate energy minimization via graph cuts. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1222–1239. Google Scholar
- Buades, A., Coll, B., & Morel, J. (2005a). a non-local algorithm for image denoising. In Proc. IEEE computer society conf. CVPR (Vol. 2, pp. 60–65). Google Scholar
- Buades, A., Coll, B., & Morel, J. M. (2009). Non-local means denoising. Image Processing on Line http://www.ipol.im/pub/algo/bcm_non_local_means_denoising/.
- Cho, T., Avidan, S., & Freeman, W. (2009). The patch transform. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1489–1501. Google Scholar
- Comaniciu, D., Ramesh, V., & Meer, O. (2003). Kernel-based object tracking. IEEE Transactions on Pattern Analysis and Machine Intelligence, 564–577. Google Scholar
- Dabov, K., Foi, A., Katkovnik, V., & Egiazarian, K. (2008). A nonlocal and shape-adaptive transform-domain collaborative filtering. In Int. workshop on local and non-local approximation in image processing, LNLA. Google Scholar
- Deledalle, C., Denis, L., & Tupin, F. (2009a). Débruitage non-local itératif fondé sur un critère de similarité probabiliste. In Proceedings of GRETSI Dijon, France, September 2009. Google Scholar
- Deledalle, C., Denis, L., & Tupin, F. (2011a). NL-InSAR: non-local interferogram estimation. IEEE Transaction on Geoscience and Remote Sensing, 49, 4. Google Scholar
- Deledalle, C. A., Duval, V., & Salmon, J. (2011b). Non-local methods with shape-adaptive patches (NLM-SAP). Journal of Mathematical Imaging and Vision, 1–18. Google Scholar
- Deledalle, C. A., Tupin, F., & Denis, L. (2011c). Patch similarity under non-Gaussian noise. In Image processing (ICIP), 18th IEEE international conference on. New York: IEEE. Google Scholar
- Efros, A., & Freeman, W. (2001). Image quilting for texture synthesis and transfer. In Proceedings of the 28th annual conference on computer graphics and interactive techniques (pp. 341–346). New York: ACM. Google Scholar
- Freeman, W., Jones, T., & Pasztor, E. (2002). Example-based super-resolution. IEEE Computer Graphics and Applications, 56–65. Google Scholar
- Ishikawa, H. (2003). Exact optimization for Markov random fields with convex priors. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1333–1336. Google Scholar
- Kay, S. (1998). Fundamentals of statistical signal processing. Volume 2: Detection theory. New York: Prentice Hall. Google Scholar
- Kendall, M., & Stuart, A. (1979). The advanced theory of statistics. Vol. 2: Inference and relationship. London: Charles Griffin and Co, Ltd. Google Scholar
- Kervrann, C., Boulanger, J., & Coupé, P. (2007). Bayesian non-local means filter, image redundancy and adaptive dictionaries for noise removal. In Proceedings of the 1st international conference on scale space and variational methods in computer vision (pp. 520–532). Berlin: Springer. CrossRefGoogle Scholar
- Mairal, J., Bach, F., Ponce, J., Sapiro, G., & Zisserman, A. (2009). Non-local sparse models for image restoration. In ICCV. Google Scholar
- Mäkitalo, M., Foi, A., Fevralev, D., & Lukin, V. (2010). Denoising of single-look SAR images based on variance stabilization and nonlocal filters. In Proc. int. conf. math. meth. electromagn. th., MMET 2010, Kiev, Ukraine. Google Scholar
- Minka, T. (1998). Bayesian inference, entropy, and the multinomial distribution (Tech. rep.), CMU. Google Scholar
- Minka, T. (2000). Distance measures as prior probabilities (Tech. rep.), CMU. Google Scholar
- Parrilli, S., Poderico, M., Angelino, C., Scarpa, G., & Verdoliva, L. (2010). A nonlocal approach for SAR image denoising. In Geoscience and remote sensing symposium (IGARSS) (pp. 726–729). New York: IEEE. Google Scholar
- Seeger, M. (2002). Covariance kernels from Bayesian generative models. In Advances in neural information processing systems 14: proceedings of the 2001 conference (p. 905). Cambridge: MIT Press. Google Scholar
- Teuber, T., & Lang, A. (2011). A new similarity measure for nonlocal filtering in the presence of multiplicative noise (Preprint). University of Kaiserslautern. Google Scholar
- Varma, M., & Zisserman, A. (2003). Texture classification: are filter banks necessary. In Computer vision and pattern recognition, 2003. Proceedings. 2003 IEEE computer society conference on (Vol. 2, pp. II–691). New York: IEEE. Google Scholar
- Yianilos, P. (1995). Metric learning via normal mixtures (Tech. rep.). NEC Research Institute, Princeton, NJ. Google Scholar