Chair: Jan Biemond, Delft University of Technology (THE NETHERLANDS)
Dilip G. Warrier, John Hopkins University (USA)
Uday B. Desai, Indian Institute of Technology (INDIA)
Any imaging system has to deal with the problem of degradation of images due to blurring and noise. In this paper, we consider the problem of blur identification, in the presence of noise. We use a stochastic model for the blur matrix. Then, we use the mean field approximation technique which enables us to obtain a closed-form expression for the mean values of the blur elements. This technique is proven to be versatile enough to handle wide varieties of blur matrices. Simulation results for estimation of the blur matrix using this formulation have also been presented.
G. Faye Boudreaux-Bartels, University of Rhode Island
Antonio H. Costa, University of Massachusetts- Dartmouth (USA)
This paper advances a novel and relatively simple scheme for designing two-dimensional (2-D) filters with zero phase. The proposed filters are obtained by a nonlinear mapping of a one-dimensional (1-D) magnitude response into a 2-D filter. By appropriately constraining the filter parameters, a wide variety of passband iso-contour shapes can be generated in the frequency-frequency plane, e.g., tilted or untilted ellipses, circles, diamonds, parallel strips at arbitrary angles, crosses, and snowflakes. Simple equations for designing the filter's parameters that meet or exceed user specifications are given for the special cases when the 1-D prototype is the magnitude response of a Gaussian, Butterworth, Chebyshev, or inverse Chebyshev filter. A scheme for designing fan filters with arbitrary angle is also provided.
Yu-Li You, University of Minnesota (USA)
M. Kaveh, University of Minnesota (USA)
This paper presents shift-adaptive blind image restoration algorithms which can deal with realistic shift-variant blurs and which integrate the usually separate tasks of blur identification and image restoration. The key to success is the effective utilization of the piecewise smoothness of both the image and the PSF to compensate for the severe lack of information in this type of problems. This is achieved through regularization of the image and the PSF by anisotropic diffusion which has the property that smoothing is allowed only in the direction of edges.
X. Yu, Washington State University
C.S. Hsu, Washington State University
R.H. Bamberger, Washington State University
S. J. Reeves, Auburn University (USA)
This paper addresses the use of H(infinity) filtering to deconvolution, in particular, to the problem of image restoration. The proposed H(infinity) deconvolution filter has some advantages in the image restoration such as it can deal with unknown boundary problem and spatially varying blurs. In this paper, H(infinity) filter is compared with the inverse Wiener filter and a regularized restoration. The experimental results show that the H(infinity) filter deals with the unknown boundary problem better than the Wiener filter. Compared with the regularization method, it gives a sharper restored image, especially, when the original image contains many details.
Y. Zhang, Beckman Institute
T.S. Huang, Beckman Institute
R.J. Adrian, University of Illinois at Urbana-Champaign (USA)
A new approach of image processing is proposed to enhance the spatial resolution of fluid flow field measurement. The local estimate velocity is first found in a sub-image region. The particles in the region are then located individually. The corresponding particles are matched using the estimate velocity and the particle locations in conjunction, making it possible to produce more velocity measurements from the smallest PIV measurement sub-region. The proposed method has been proved successful by testing the experimental result from Urushihara et. al [6]