Home HomeJitendra K. Tugnait, Auburn University (U.S.A.)
This paper is concerned with the problem of estimating the multichannel impulse response function of a 2-D multiple-input multiple-output (MIMO) system given only the measurements of the vector output of the system. Such models arise in a variety of situations such as color images (textures), or image data from multiple frequency bands, multiple sensors or multiple time frames. We extend the approach of Tugnait(1994) (which deals with SISO 2-D systems) to MIMO 2-D systems. The paper is focused on certain theoretical aspects of the problem: estimation criteria, existence of a solution, and parameter identifiability. An iterative, inverse filter criteria based approach is developed using the third-order and/or fourth-order normalized cumulants of the inverse filtered data at zero-lag. The approach is input-iterative, i.e., the input sequences are extracted and removed one-by-one. The matrix impulse response is then obtained by cross-correlating the extracted inputs with the observed outputs.
ic973029.pdf ic973029.pdf
TOP
Stéphane Coulombe, INRS-Télécommunications (Canada)
Eric Dubois, INRS-Télécommunications (Canada)
This article addresses the problem of designing two-channel near-perfect-reconstruction filter banks over multidimensional lattices. First, a cosine-modulated filter structure having arbitrary spatial shift and phase parameters is considered. The use of this structure leads to many possible two-channel multirate systems. The perfect reconstruction conditions are studied and appropriate choices for the parameters of the cosine-modulated structure are obtained. A simple but efficient unconstrained procedure for designing (possibly linear phase) near-perfect-reconstruction filter banks having good frequency responses (with arbitrary shapes) is proposed. A 2D design example is presented. The filter banks obtained can be used in transmultiplexers as well as in subband coders.
ic973033.pdf ic973033.pdf TOP
Amir Asif, Ghulam Ishaq Khan Institute (Pakistan)
José M.F. Moura, Carnegie Mellon University (U.S.A.)
We develop computationally fast and storage efficient implementations for the Kalman-Bucy filter (KBf) for data assimilation problems with large time varying multidimensional fields. We refer to them as the block KBf (bKBf) and the localized block KBf (lbKBf). For fields defined on a 2D lattice of linear dimension $I$, the bKBf reduces the computational complexity of the KBf by O(I). The lbKBf saves further on computations by a factor of I and decreases the storage requirements by O(I). We illustrate the lbKBf in assimilating satellite measurements in physical oceanography, presenting simulations for an equatorial beta plane.
ic973037.pdf ic973037.pdf TOP
Krishna Ratakonda, University of Illinois, Urbana. (U.S.A.)
Narendra Ahuja, University of Illinois, Urbana. (U.S.A.)
Inorder to apply a multi-dimensional linear transform over an arbitrarily shaped support, the usual practice is to fill out the support to a hypercube by zero padding. This does not however yield a satisfactory definition for transforms in two or more dimensions. The problem that we tackle is: how do we redefine the transform over an arbitrary shaped region suited to a given application? We present a novel iterative approach to define any multi-dimensional linear transform over an arbitrary shape given that we know its definition over a hyper-cube. The proposed solution is (1) extensible to all possible shapes of support (whether connected or unconnected) (2) adaptable to the needs of a particular application. We also present results for the Fourier Transform, for a specific adaptation of the general definition of the transform which is suitable to compression or segmentation algorithms.
ic973041.pdf ic973041.pdf TOPRiccardo Bernardini, EPFL (Switzerland)
Cosine modulated filter banks are a well-known signal processing tool whose applicative field ranges from coding, to filtering, to spectral estimation. Because of their peculiar structure (the impulse responses are obtained by modulating a prototype window with trigonometric functions) they are easy to design and have a low computation complexity. Their continuous-time counterpart, local cosine bases, play an important role in the construction of Lemarié-Meyer wavelets. We propose a unified approach to both discrete and continuous time cosine modulated filter banks. The resulting theory offers a single general framework that makes clear the deep similarity between the two cases.
ic973045.pdf ic973045.pdf TOP
Adrian G. Borş, University of Thessaloniki (Greece)
Ioannis Pitas, University of Thessaloniki (Greece)
William Puech, TIMC-IMAG Lab. (France)
Jean-Marc Chassery, TIMC-IMAG Lab. (France)
A set of monocular images of a curved painting is taken from different viewpoints around its curved surface. After deriving the surface localization in the camera coordinate system we backproject the image on the curved surface and we flatten it. We analyze the perspective distortions of the scene in the case when it is mapped on a cylindrical surface. Based on the result of this analysis we derive the necessary number of views in order to represent the entire scene depicted on a cylindrical surface. We employ a matching-based mosaicing method for reconstructing the scene from the curved surface. The proposed method is appropriate to be used for painting reconstruction.
ic973049.pdf ic973049.pdf TOP
Benoît Duc, EPFL (Switzerland)
Stefan Fischer, EPFL (Switzerland)
Josef Bigün, EPFL (Switzerland)
This paper investigates the application of statistical pattern recognition methods in the framework of the Dynamic Link Matching approach. This method describes objects by means of local frequency information on nodes of a sparse grid. Matching of an input image with a reference is achieved by displacement and deformation of the grid. This method is applied here to the authentication of human faces in a cooperative scenario where candidates claim an identity that is to be checked. The matching error is not powerful enough to provide satisfying results in this case. We introduce an automatic weighting of the nodes according to their significance. Results show that for regular grids, this weighting leads to a significant improvement of the performance.
ic973053.pdf ic973053.pdf TOPErhard Schubert, ZESS (Germany)
The acquisition and measurement of two- and three-dimensional contours of objects are important tasks in modern production processes and quality control. Specially, the nontactile methods like optical triangulation and the digital image processing get more and more importance. The color image processing in combination with a color coded illumination could be used to realize new methods of nontactile 3D-object ranging. Two of these methods are the color-coded triangulation and the color-coded phase-shift method. We use a combination of these two methods to realize a fast 3D-object ranging with unambiguous results. The color-coded phase-shift method is able to reach a good spatial resolution, but the measured range values are ambiguous. Using the color-coded triangulation an unambiguous three-dimensional image could be achieved, but compared to the color-coded phase-shift method, the spatial resolution is poor. Since both methods are able to generate a 3D-object description by processing only a single RGB-image, it is possible to combine these two methods.
ic973057.pdf ic973057.pdf TOPMohammed Abdul-Hameed, Elect. & Comm. Dept., Baghdad University (Iraq)
The optimum computation of multi-dimensional (multi-D) image moments is presented in this paper. The developed algorithm is designed for the general case, specifically, to an arbitrary moment order R and to dimension d. The properties of the algorithm makes it best suited for obtaining the well known 2-D Zernike moments when they are computed through their relation to ordinary moments. Computational complexity model shows that the proposed algorithm takes only (NR+N)(N+R+1) additions with a negligible amount of multiplications, when an N-sized image is used to generate 2-D ordinary moments up to the order R. While the speed improvement of obtaining Zernike moments is of the order O(R) with respect to direct computation through Zernike polynomials. The regular structure of the processing elements and the minimum no. of operations of the algorithm makes it best suited for hardware and software implementations.
ic973061.pdf TOP
Vinod Chandran, Queensland University of Technology (Australia)
Stefan Slomka, Queensland University of Technology (Australia)
Megan Gollogly, Queensland University of Technology (Australia)
Steve Elgar, Washington State University (U.S.A.)
Features derived from the trispectra of DFT magnitude slices are used for multi-font digit recognition. These features are insensitive to translation, rotation, or scaling of the input. They are also robust to noise. Classification accuracy tests were conducted on a common data base of 256 x 256 pixel bilevel images of digits in 9 fonts. Randomly rotated and translated noisy versions were used for training and testing. The results indicate that the trispectral features are better than moment invariants and affine moment invariants. They achieve a classification accuracy of 95% compared to about 81% for Hu's moment invariants and 39% for Flusser/Suk affine moment invariants on the same data in the presence of 1% impulse noise using a 1-NN classifier. A multilayer perceptron with no normalization for rotations and translations yields 34% accuracy on 16 x 16 pixel low-pass filtered and decimated versions of the same data.
ic973065.pdf ic973065.pdf TOP
Metin Nafi Gürcan, Bilkent University (Turkey)
Yasemin Yardmc, Bilkent University (Turkey)
A. Enis Çetin, Bilkent University (Turkey)
Rashid Ansari, Bilkent University (Turkey)
In this paper, computer-aided detection and enhancement of microcalcifications in mammogram images are considered. The mammogram image is first decomposed into subimages using a 'subband' decomposition filter bank which uses nonlinear filters. A suitably identified subimage is divided into overlapping square regions in which skewness and kurtosis as measures of the asymmetry and impulsiveness of the distribution are estimated. All regions with high positive skewness and kurtosis are marked as a regions of interest. Next, an outlier labeling method is used to find the locations of microcalcifications in these regions. An enhanced mammogram image is also obtained by emphasizing the microcalcification locations. Linear and nonlinear subband decomposition structures are compared in terms of their effectiveness in finding microcalcificated regions and their computational complexity. Simulation studies based on real mammogram images are presented.
ic973069.pdf ic973069.pdf TOP
Ilangko Balasingham, NTNU, Trondheim (Norway)
Tor Audun Ramstad, NTNU, Trondheim (Norway)
John Markus Lervik, NTNU, Trondheim (Norway)
The performance of subband image coders depends on proper choice of filter banks. Although odd length filters in the filter banks produce waveform type artifacts, this can be alleviated by enforcing a smooth interpolation property to the synthesis lowpass filter. Evaluation of even, odd, and combinations of even and odd length filters in tree-structured filter banks, where the filter coefficients are obtained by optimizing for coding gain at each stage, is done for image coding purposes. Favorable results are obtained when a combination of odd and even length filters are used.
ic973073.pdf ic973073.pdf TOP
Amy S. Rosenthal, Bell Labs (U.S.A.)
Jianying Hu, Bell Labs (U.S.A.)
Michael K. Brown, Bell Labs (U.S.A.)
We introduce a new method for size and orientation normalization of unconstrained handwritten words based on the Hough transform. A modified Hough transform is applied to extremum points along the Y coordinate to extract parallel lines corresponding to the boundary lines separating different vertical zones of the handwritten word. One dimensional Gaussian smoothing with variable variance is applied in the Hough space to alleviate the problems caused by the large variation in natural handwriting and the sparseness of extremum points. The method has been tested with and incorporated into an HMM based writer-independent, unconstrained on-line handwriting recognition system and a 25% error rate reduction has been achieved.
ic973077.pdf ic973077.pdf TOP
Tetsunori Kobayashi, Waseda University (Japan)
Satoshi Haruyama, Waseda University (Japan)
A new pattern matching method, Partly-Hidden Markov model, is proposed for gesture recognition. Hidden Markov Model, which is widely used for the time series pattern recognition, can deal with only piecewise stationary stochastic process. We solved this problem by introducing the modified second order Markov Model, in which the first state is hidden and the second one is observable. As the results of 6 sign-language recognition test, the error rate was improved by 73% compared with normal HMM.
ic973081.pdf TOP