Nonlocal-means (NL-means) is an image denoising method that replaces each pixel by a weighted average of all the pixels in the image. Unfortunately, the method requires the computation of the weighting terms for all possible pairs of pixels, making it computationally expensive. Some short-cuts assign a weight of zero to any pixel pairs whose neighbourhood… (More)
In this paper we present a novel single-frame image zooming technique based on so-called " self-examples ". Our method combines the ideas of fractal-based image zooming, example-based zooming, and nonlocal-means image denoising in a consistent and improved framework. In Bayesian terms, this example-based zooming technique targets the MMSE estimate by… (More)
We introduce a novel super-resolution scheme for multi-frame image sequences. Our method is closely associated with the recently developed " non-local-means denoising filter ". In the proposed algorithm, no explicit motion estimation is performed, unlike in many other methods. Our results are comparable, if not superior, to many existing approaches,… (More)
This paper presents a general PDE-framework for registration of contrast enhanced images. The approach directly applies the idea of separating the contrast enhancement term from the images in the regularization terms. In our formulation, we stay consistent with existing non-parametric image registration techniques, however, we carry an additional contrast… (More)
We show how fractal image coding can be viewed and generalized in terms of the method of projections onto convex sets (POCS). In this approach, the fractal code defines a set of spatial domain similarity constraints. We also show how such a reformulation in terms of POCS allows additional contraints to be imposed during fractal image decoding. Two… (More)
The goal of this paper is to present a novel recipe for de-formable image registration under varying illumination, as a natural extension of the demons algorithm. This generalization is derived directly from the optical-flow constraints in a variational formulation. Furthermore , our approach provides a new mathematical interpretation of the demons… (More)
—The method of nonlocal-means is a powerful method for image denoising, but comes with a heavy computational price. The computation can be accelerated using the Fast Fourier Transform (FFT), yielding speed gains by a factor of 8 or more. Experiments show the benefits of the FFT in this context.
The goal of this paper is to present a new recipe for the fractal image decoding process. In this paper, we explain how fractal-based methods can be internally combined with regularization schemes, e.g., Tikhonov, total variation (TV), or hard-constrained regularization. Although the regularization procedure is very common in context of algebraic image… (More)