HomeChair: Edward J. Delp, Purdue University, USA
HomeGonzalo R Arce, University of Delaware (U.S.A.)
Weighted median filters (smoothers) have been shown to be analogous to normalize FIR linear filters constrained to have only positive weights. In this paper, it is shown that much like the mean is generalized to the rich class of linear FIR filters, the median can be generalized to a richer class of weighted median (WM) filters admitting positive and negative weights. The generalization follows naturaly and is surprisingly simple. In order to analyze and design this class of WM filters, a new threshold decomposition theory admitting real-valued input signals is developed which, in turn, is used to develop fast adaptive algorithms to optimally design the real-valued filter coefficients. The new WM filter formulation leads to significantlymore powerful estimators capable of effectively addressing a number of fundamental problems in signal processing which could not adequately be addressed by prior WM filter (smoother) structures.
ic981261.pdf (From Postscript)
TOP
Nir Sochen, Technion - Institute of Technology (Israel)
Yehoshua Y Zeevi, Technion - Institute of Technology (Israel)
Algorithms for resolution enhancement are needed in various applications of image processing and communication suchas compression, and HDTV, in which the enhancement of low resolution images acquired by CCD--based electronic color cameras is required. We develop a geometrical algorithm, based on diffusion processes, which are used both for smoothing of the enlarged image, when ``time"" is flowing forward, and for enhancement. The latter is accomplished by allowing the ``time"" to flow backwards, i.e. ``solving"" an inverse diffusion problem which is ill posed. In order to stabilize the flow as well as to enhance important features (e.g. edges) on the expense of less important image domains, we use a modified Beltrami diffusion equation.
ic981504.pdf (From Postscript) TOP
David G Sheppard, University of Arizona (U.S.A.)
Bobby R Hunt, University of Arizona (U.S.A.)
Michael W Marcellin, University of Arizona (U.S.A.)
Algorithms for image recovery with super-resolution from sequences of short-exposure images are presented in this paper. Both deconvolution from wavefront sensing (DWFS) and blind deconvolution are explored. A multiframe algorithm is presented for DWFS which is based on maximum a posteriori (MAP) formulation. A multiframe blind deconvolution algorithm is presented based on a maximum likelihood formulation with strict constraints incorporated using nonlinear reparameterizations. Quantitative simulation of imaging through atmospheric turbulence and wavefront sensing are used to demonstrate the super-resolution performance of the algorithms.
ic981418.pdf (From Postscript) TOP
Hung-Ta Pai, University of Texas, Austin (U.S.A.)
John W Havlicek, University of Texas, Austin (U.S.A.)
Alan C Bovik, University of Texas, Austin (U.S.A.)
We have previously developed an algorithm and sufficient conditions for exact blind image restoration. In this paper, we use the resultant matrix theorem and techniques of algebraic geometry to prove that the sufficient conditions hold generically given three blurred versions of the same image and some restrictions on the size of the original image. Moreover, the extension to multichannel blind n-dimensional signal restoration is described.
ic982141.pdf (From Postscript) TOP
Robinson Piramuthu, University of Michigan (U.S.A.)
Alfred O. Hero III, University of Michigan (U.S.A.)
A method for incorporating anatomical MRI boundary side information into penalized maximum likelihood (PML) Emission Computed Tomography (ECT) image reconstructions using a set of averaged Gibbs weights was proposed in an earlier paper. A quadratic penalty based on Gibbs weights was used to enforce smoothness constraints everywhere in the image except across the estimated boundary of ROI. In this methodology, a limiting form of the posterior distribution of the MRI boundary parameters was used to average the Gibbs weights obtained using other techniques. We show an improvement in performance when the variance of boundary estimates from the MRI data becomes significant. In this paper, we present the empirical performance analysis of the proposed method of averaged Gibbs weights.
ic982211.pdf (From Postscript) TOP
Robert D. Nowak, Michigan State University (U.S.A.)
Michael J Thul, University of Kaiserslautern (Germany)
This paper studies linear shift-invariant inverse problems arising in photon-limited imaging. The problem we consider is the recovery of an intensity image from a distorted version degraded with Poisson noise. This problem arises in medical and astronomical imaging. It is shown that the wavelet-vaguelette decomposition (WVD) can provide much better estimates of the underlying intensity compared to classical frequency domain methods. The paper combines recently developed wavelet-based filtering techniques for photon imaging with new results in WVD methods for inverse problems. Furthermore, we show that the WVD can be interpreted as a prefiltered wavelet transform, and that it can be very efficiently computed. The new method is applied to nuclear medicine imaging.
ic982118.pdf (From Postscript) TOP
Jr-Jen Huang, Purdue University (U.S.A.)
Edward J. Coyle, Purdue University (U.S.A.)
The present approach to the MAE-based design of stack filters for image restoration does not always produce the desired visual result. Thus, in this paper, a new stack filter design algorithm is developed. It is based upon a Weighted Mean Absolute Error (WMAE) criterion instead of the traditional MAE criterion, which assigns the same weights to all errors. The weights in this WMAE criterion are designed with the aid of the Visible Differences Predictor (VDP), which can estimate the sensitivity of the human visual system to changes in images. Experiments with this WMAE approach show that the stack filters it produces perform significantly better in image processing applications than those designed with the MAE approach.
ic982526.pdf (From Postscript) TOP
Richard Rau, Georgia Institute of Technology (U.S.A.)
James H. McClellan, Georgia Institute of Technology (U.S.A.)
This paper presents a theoretical analysis of the structure of wide angle, ultra-wideband SAR images formed by a constant integration angle backprojection image former. It is shown that the effects of the image former can be modeled as a filtering operation on the original data. Furthermore, SAR images for different squint angles can be obtained from the original images by directional filtering. As a result, it will be shown that perfect reconstructing directional filterbanks can be used as a unitary transform between SAR images and a 3-D representation containing additional aspect-angle information. It will be demonstrated how this new representation can be used to enhance targets.
ic981809.pdf (From Postscript) TOP