Image Denoising

One of the fundamental challenges in the field of image processing and computer vision is image denoising, where the underlying goal is to estimate the original image by suppressing noise from a noise-contaminated version of the image. Image noise may be caused by different intrinsic (i.e., sensor) and extrinsic (i.e., environment) conditions which are often not possible to avoid in practical situations. Therefore, image denoising plays an important role in a wide range of applications such as image restoration, visual tracking, image registration, image segmentation, and image classification, where obtaining the original image content is crucial for strong performance. While many algorithms have been proposed for the purpose of image denoising, the problem of image noise suppression remains an open challenge, especially in situations where the images are acquired under poor conditions where the noise level is very high.

In the VIP lab, we investigate an alternative approach to the problem of image denoising based on data-adaptive stochastic optimization via Markov-Chain Monte Carlo sampling. By formulating the problem as a Bayesian optimization problem and taking a nonparametric stochastic strategy to solving this problem, such a Markov-Chain Monte Carlo denoising (MCMCD) strategy dynamically adapts to the underlying image and noise statistics in a flexible manner to provide high denoising performance while maintaining relatively low computational complexity.

Examples of the results produced using the MCMCD strategy are shown below:

Image of hill, with a lot of image noise
Image of hill after image denoising has taken place
 
Image of Barbara before image denoising
Image of Barbara after image denoising
 

 Related people

Directors

Alexander Wong, David A. Clausi, Paul Fieguth

Students

Edward Li

Alumni

Akshaya Mishra, Fu Jin, Wen Zhang

Related research areas

Image Segmentation/Classification

Related publications

Journal articles

Wong, A., and J. Scharcanski, "Monte Carlo Despeckling of Transrectal Ultrasound (TRUS) Images of the Prostate",Digital Signal Processing, Accepted. Details

Xu, L., J. M. Shafiee, A. Wong, and D. A. Clausi, "Fully-Connected Continuous Conditional Random Field With Stochastic Cliques for Dark-spot Detection In SAR Imagerypdf", IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, February, Accepted. Details

Li, F.L. XuA. Wong, and D. A. Clausi, "QMCTLS: Quasi Monte Carlo Texture Likelihood Sampling for Despeckling of Complex Polarimetric SAR Imagespdf", IEEE Geoscience and Remote Sensing Letters, vol. 12, issue 7, February, 2015. Details

Xu, L.F. LiA. Wong, and D. A. Clausi, "Hyperspectral Image Denoising Using a Spatial–Spectral Monte Carlo Sampling Approachpdf", IEEE Journal of Selected Topics on Applied Earth Observations and Remote Sensing, vol. 8, issue 6: IEEE, 2015. Details

Xu, L.A. WongF. Li, and D. A. Clausi, "Intrinsic Representation of Hyperspectral Imagery for Unsupervised Feature Extractionpdf", IEEE Transactions on Geosciences and Remote Sensing, vol. 54, issue 2: IEEE, 2015. Details

Wang, C., L. XuD. A. Clausi, and A. Wong, "A Bayesian Joint Decorrelation and Despeckling approach for speckle reduction of SAR Images", Vision Letters, vol. 1, issue 1, 2015. Details

Xu, L., J. Li, Y. Shu, and J. Peng, "SAR Image Denoising via Clustering-Based Principal Component Analysispdf", IEEE Transactions on Geoscience and Remote Sensing, vol. 52, issue 11: IEEE, pp. 6858-6869, 2014. Details

Lui, D.A. Cameron, A. Modhafar, D. Cho, and A. Wong, "Low-dose computed tomography via Spatially-adaptive Monte Carlo reconstruction", Computerized Medical Imaging and Graphics, vol. 37, issue 7-8, pp. 438 - 449, October, 2013. Details

Cameron, A.D. LuiA. BoroomandJ. GlaisterA. Wong, and K. Bizheva, "Stochastic speckle noise compensation in optical coherence tomography using non-stationary spline-based speckle noise modelling", Biomedical Optics Express, vol. 4, no. 9, August, 2013. Details

Wong, A.A. MishraW. ZhangP. Fieguth, and D. A. Clausi, "Stochastic image denoising based on Markov-chain Monte Carlo samplingpdf", Signal Processing, vol. 91, issue 8, pp. 2112 - 2120, 2011. Details

Wong, A.A. Mishra, K. Bizheva, and D. A. Clausi, "General Bayesian estimation for speckle noise reduction in optical coherence tomography retinal imagerypdf", Optics Express, vol. 18, no. 8, pp. 8338–8352, 2010. Details

Wong, A., "Adaptive bilateral filtering of image signals using local phase characteristicspdf", Signal Processing, vol. 88, no. 6, pp. 1615–1619, 2008. Details

Conference papers

Li, E.M. J. ShafieeF. Kazemzadeh, and A. Wong, "Sparse Reconstruction of Compressed Sensing Multi-spectral Data using Cross-Spectral Multi-layered Conditional Random Field Modelpdf", SPIE optics and photonics 2015, San Diego, SPIE , May, Accepted. Details

Leigh, A., A. WongD. A. Clausi, and P. Fieguth, "Comprehensive analysis on the effects of noise estimation strategies on image noise artifact suppression performance", IEEE International Symposium on Multimedia, 2011. Details

Eichel, J. A., D. Lee, A. WongP. FieguthD. A. Clausi, and K. Bizheva, "Quantitative comparison of despeckling and frame averaging approaches to processing retinal OCT tomograms", SPIE Photonics West (BiOS), 2011. Details