Multi-scale MAP estimation of high-resolution images

Authors:
Shubin Zhao
Affiliations:
Jiangsu Automation Research Institute, Lianyungang, Jiangsu, China
Venue:
PSIVT'06 Proceedings of the First Pacific Rim conference on Advances in Image and Video Technology
Year:
2006

Citing 9
Cited 0

Nonlinear total variation based noise removal algorithms

Proceedings of the eleventh annual international conference of the Center for Nonlinear Studies on Experimental mathematics : computational issues in nonlinear science: computational issues in nonlinear science
Limits on Super-Resolution and How to Break Them

IEEE Transactions on Pattern Analysis and Machine Intelligence
Hallucinating Faces

FG '00 Proceedings of the Fourth IEEE International Conference on Automatic Face and Gesture Recognition 2000
Wavelet-based statistical signal processing using hidden Markovmodels

IEEE Transactions on Signal Processing
Extraction of high-resolution frames from video sequences

IEEE Transactions on Image Processing
Joint MAP registration and high-resolution image estimation using a sequence of undersampled images

IEEE Transactions on Image Processing
Restoration of a single superresolution image from several blurred, noisy, and undersampled measured images

IEEE Transactions on Image Processing
Wavelet domain image restoration with adaptive edge-preserving regularization

IEEE Transactions on Image Processing
A computationally efficient superresolution image reconstruction algorithm

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, a multi-scale MAP algorithm for image super-resolution is proposed. It is well known that Reconstructing high-resolution(HR) images from multiple low-resolution(LR) images or a single one is an ill-posed problem. The main challenge is how to preserve edges in images while reducing noise. According to Bayesian approaches, which are popular and widely researched, solving this kind of problems is introducing prior knowledge about HR images as constraints and obtaining good HR images in some sense. In this paper, wavelet-domain prior distributions are concisely analyzed. And then, by introducing wavelet-domain Hidden Markov Tree-structured model(HMT) which accurately characterizes the statistics of most real-world images, reconstruction of HR images is reformulated as a multi-scale MAP estimation problem. For justification of this formulation, HMT is interpreted in the regularization framework, concisely and clearly. Experimental results are presented for assessment.