The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
A hidden-Markov-model (HMM)-based system for font-independent spotting of user-specified keywords in a scanned image is described. Word bounding boxes of potential keywords are extracted from the image using a morphology-based preprocessor. Feature vectors based on the external shape and internal structure of the word are computed over vertical columns of pixels in a word bounding box. For each user-specified...
Several objective functions for vector field segmentation are presented. Y. G. Leclerc's (1989) MRF (Markov random field) model is extended by the addition of information-theoretic penalties for regions and distinct means. Standard methods of signal detection and estimation are used to develop a theoretical performance analysis which quantitatively predicts the performance at realistic noise levels...
By making use of the Hough and the contour sequence matching technique, the authors suggest a novel approach for solving the problem of the large memory requirement of the Hough space. For the conventional Hough algorithm it is necessary to take a four-dimensional space for the recognition of an irregular pattern, while in the proposed algorithm the Hough space is replaced with a two-dimensional domain...
An empirical measure for the selection of the edge-enhancement Gaussian filter is developed. The Gaussian filter is specified by its standard deviation sigma ; the filter's spatial support is a function of sigma . An estimation procedure for sigma using Fourier analysis is described. The measure is easy to implement and is based totally on the image at hand. Experimental results suggest that this...
A novel approach for unsupervised image segmentation is described. This approach makes use of a Gaussian pyramid as multiresolution decomposition to analyze images. Compound random fields are used to model images at each resolution. The hierarchical image model is formed by a Strauss process in the lower level and a set of white Gaussian random fields in the upper level. This basic image model is...
In this scheme, each texture is modeled as one HMM. Thus, if there are M different textures present in an image, there are M distinct HMMs to be found and trained. Consequently, the unsupervised texture segmentation problem becomes an HMM-based problem, where the appropriate number of HMMs, the associated model parameters, and the discrimination among the HMMs are the foci of the scheme. The scheme...
An algorithm for image interpolation that takes into account both the motion of the camera and the motion of the imaged objects is described. First the estimated global motion parameters are used to compensate for the camera motion. The missing images are then interpolated by using the local motion field. Finally, these images are reconstructed in their correct dimension, considering the global motion...
A novel approximation of Euclidean distance in Z/sup 2/ is proposed, and a novel algorithm for the computation of Voronoi tessellation and Delauney triangulation is presented based on this approximation. The proposed method has low computational complexity (of order O(1/N)) and allows parallel implementation. Mathematical morphology is used to implement the Voronoi tessellation and the Delauney triangulation...
The authors propose a general formulation for adaptive, maximum a posteriori probability (MAP) segmentation of image sequences on the basis of interframe displacement and gray level information. The segmentation classifies pixel sites to independently moving objects in the scene. In this formulation, two methods for characterizing the conditional probability distribution of the data given the segmentation...
An algorithm for designing optimal Gabor filters is presented. The algorithm assumes that an image contains two different textures and that prototype samples of the desired textures are given. It uses a decision-theoretic framework, based on modeling a Gabor-filter output as a Rician distribution, for designing optimal filters. To gain more robust results, a multiple-filter segmentation scheme is...
A framework for multiscale stochastic modeling was introduced (K.C. Chou et al., 1989) based on coarse-to-fine scale-recursive dynamics defined on trees. This model class has some attractive characteristics which lead to extremely efficient, statistically optimal signal and image processing algorithms. In the present work, the authors describe how 1-D Markov processes and 2-D Markov random fields...
The estimation properties of morphological filters are considered in terms of edge localization and grey level contrast preservation in two-dimensional spaces. It is shown that, at least in practice, a compromise between these two characteristics has to be made. In a first step, morphological filters are compared with linear and median filters. It is concluded that an efficient edge localization can...
The authors describe some iterative segmentation algorithms that combine statistical constraints represented in Markov random field models with deterministic constraints imposed by morphological operations. The goal is to produce segmentations that have high probability according to the Markov model and are smooth in the sense of being morphologically open and/or closed. The authors first present...
An attempt is made to apply a novel multidirectional resolution analysis to obtain boundaries between areas with different grades of 'roughness'. To this end, the image texture is parametrized with respect to direction, based on the model of isotropic 2-D fractional Brownian motion (2-D FBM). Some theoretical aspects of this model have not been clarified yet. The authors first provide some of the...
It is shown how linear combinations of morphological operators can be formed to alleviate the bias introduced by the individual morphological operators. Since every morphological operator has a complementary operator that is equally and oppositely biased, the authors propose averaging the complementary operators to alleviate the bias. Of the three filters formed by averaging the standard morphological...
A wavelet approach to linear inverse problems in image processing is described. The images and the operator to be inverted are expanded by wavelets and various constraints for a regularized solution are enforced through wavelet coefficients. The approach also provides a solution to an important problem in multigrid/multiresolution processing: representing an operator in different resolutions. The...
The authors describe symmetric convolution and its use for the nonexpansive implementation of multirate filter banks for images. Symmetric convolution is a formalized approach to convolving symmetric FIR (finite impulse response) filters with symmetrically extended data. It is efficient because the discrete sine and cosine transforms can be used to perform the convolution as a transform-domain multiplication...
The use of 2-D decimated median filters (DMFs) for the removal of impulsive noise and the preservation of edges is proposed. The 2-D DMF preserves edges and corners better than the median filter because its root signal set contains corners as well as edges. Its impulse noise removal capability is comparable with that of the median filter while its performance near edges is superior. The computational...
A method is developed for the synthesis of a nonlinear adaptive filter based on solutions to the inhomogeneous diffusion equation. The approach is based on the specification of the first derivative of the signal in time (scale). A general solution is derived and is then specialized to the scale invariance case, in which the diffusion coefficient is shown to be the gradient inverse. A novel discrete...
The authors present the results of an extensive study investigating the applicability of different existing nonlinear adaptive filtering methods, as well as a new speckle-model based quadratic Volterra filter (QVF), to solve the problem of smoothing speckle noise in digital images while preserving important edge information. The QVF consists of a linear part, mainly responsible for noise smoothing,...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.