Bayesian multi frame super resolution

Turgay, Emre
This thesis aims at increasing the effective resolution of an image using a set of low resolution images. This process is referred to as super resolution (SR) image reconstruction in the literature. This work proposes maximum a-posteriori (MAP) based iterative reconstruction methods for this problem. The first contribution of the thesis is a novel edge preserving SR image reconstruction method. The proposed MAP based estimator uses local gradient direction and amplitude for optimal noise reduction while preserving edges. The second contribution of the thesis is a novel texture prior for maximum a posteriori (MAP) based super resolution (SR) image reconstruction. The prior is based on a multiscale compound Markov Random Field (MRF) model. Gabor filters are utilized for subband decomposition. Each subband is modeled by a compound MRF that inherits a binary texture process. The texture process at each pixel location at each subband is estimated iteratively along with the unknown high-resolution image pixels. Finally, a two stage SR method comprising a Bayesian reconstruction step followed by a restoration step is proposed. In the first stage, two MAP based SR estimators with different regularizations are employed. In the second stage, pixel-to-pixel difference between these two estimates is post-processed to restore edges and textures while eliminating noise. Experiments on synthetically generated images and real experiments on visual CCD cameras and thermal cameras demonstrate that the proposed methods are more favorable compared to state-of-the-art SR methods especially on textures and edges.


An Investigation on hyperspectral image classifiers for remote sensing
Özdemir, Okan Bilge; Çetin, Yasemin; Department of Information Systems (2013)
Hyperspectral image processing is improved by the capabilities of multispectral image processing with high spectral resolution. In this thesis, we explored hyperspectral classification with Support Vector Machines (SVM), Maximum Likelihood (ML) and KNearest Neighborhood algorithms. We analyzed the effect of training data on classification accuracy. For this purpose, we implemented three different training data selection methods; first N sample selection, randomly N sample selection and uniformly N sample se...
Privacy protection of tone-mapped HDR images using false colours
ÇİFTÇİ, Serdar; Akyüz, Ahmet Oğuz; PİNHEİRO, Antonio M. G.; Ebrahimi, Touradj (2017-12-01)
High dynamic range (HDR) imaging has been developed for improved visual representation by capturing a wide range of luminance values. Owing to its properties, HDR content might lead to a larger privacy intrusion, requiring new methods for privacy protection. Previously, false colours were proved to be effective for assuring privacy protection for low dynamic range (LDR) images. In this work, the reliability of false colours when used for privacy protection of HDR images represented by tone-mapping operators...
A comparative evaluation of super – resolution methods on color images
Erbay, Fulya; Akar, Gözde; Department of Electrical and Electronics Engineering (2011)
In this thesis, it is proposed to get the high definition color images by using super – resolution algorithms. Resolution enhancement of RGB, HSV and YIQ color domain images is presented. In this study, three solution methods are presented to improve the resolution of HSV color domain images. These solution methods are suggested to beat the color artifacts on super resolution image and decrease the computational complexity in HSV domain applications. PSNR values are measured and compared with the results of...
Cognitive aspects of image zooming
Öncül, Gaye; Tarı, Sibel; Tekman, Hasan Gürkan; Department of Cognitive Sciences (2002)
In this work, digital image zooming methods and image distortion metrics are examined in terms of a quality criterion based on image appearance. Neurophysiological and psychophysical foundations of perception driven image distortion metrics are presented. Results of a comparative study based on both perceptual distortion metrics and conventional distortion metrics for standard zooming methods are discussed. Open problems in image quality definition as well as image enlargement methods are presented and impr...
Image resolution enhancement using wavelet domain Hidden Markov Tree and coefficient sign estimation
Temizel, Alptekin (2007-01-01)
Image resolution enhancement using wavelets is a relatively new subject and many new algorithms have been proposed recently. These algorithms assume that the low resolution image is the approximation subband of a higher resolution image and attempts to estimate the unknown detail coefficients to reconstruct a high resolution image. A subset of these recent approaches utilized probabilistic models to estimate these unknown coefficients. Particularly, hidden Markov tree (HMT) based methods using Gaussian mixt...
Citation Formats
E. Turgay, “Bayesian multi frame super resolution,” Ph.D. - Doctoral Program, Middle East Technical University, 2014.