Showing papers in "Graphical Models and Image Processing in 1996"

TL;DR: A theory of fuzzy objects forn-dimensional digital spaces based on a notion of fuzzy connectedness of image elements and algorithms for extracting a specified fuzzy object and for identifying all fuzzy objects present in the image data are presented.

TL;DR: This paper approaches all three problems of reconstruction of the 3D object from the 2D scalar-valued slices obtained by the imaging system simultaneously by imposing a set of three constraints on the reconstructed surface and then deriving precise correspondence and tiling rules from these constraints.

TL;DR: A class of PDE-based algorithms suitable for image denoising and enhancement based on a curvature-controlled approach that is applicable to both salt-and-peppergray-scale noise and full-image continuous noise present in black and white images, gray-scale images, texture images, and color images.

TL;DR: A system which automatically annotates images with a set of prespecified keywords, based on supervised color classification of pixels intoprespecified classes using simple pixelwise operations is described.

TL;DR: This paper presents a classical multiscale model which consists of a label pyramid and a whole observation field, and proposes a hierarchical Markov random field model based on this classical model, which results in a relaxation algorithm with a new annealing scheme: the multitemperatureAnnealing (MTA) scheme, which consist of associating higher temperatures to higher levels in order to be less sensitive to local minima at coarser grids.

TL;DR: The Marching Lines extracts, with subpixel accuracy, the 3D lines corresponding to the intersection of two isosurfaces in a 3D grid, and ensures that each reconstructed curve is formed of a connected set of segments.

TL;DR: A novel method based on the automatic search of features that characterize a certain object class using a training set consisting of both positive and negative examples, falling under the general problems of texture recognition, texture defect detection, and shape recognition.

TL;DR: The theory of the medial axis transform for 3-D objects is developed and it is proven that path connected polyhedral solids without cavities have path connected medial axes.

TL;DR: The cellular automata methods studied present very fast convergence to fixed points, noise stability, and improvements on real images, which are features that allow them to propose them as a first level elementary image enhancement.

TL;DR: In this article, a new curve-fitting algorithm is presented, which can automatically fit a set of data points with piecewise geometrically continuous (G1) cubic Bezier curves.

TL;DR: The proposed method is capable of tracking image regions corresponding to scene objects with nonplanar and/or mildly deforming surfaces, accounting for intensity variations, and is shown to be effective with real image sequences.

TL;DR: An alternative formulation of histogram equalization is investigated that involves the application of a filter to a set of image fields, and the potential for filtering these fields differently is illustrated through the restoration of a deliberately degraded image.

TL;DR: This algorithm can automatically fit a set of data points with piecewise geometrically continuous (G1) cubic Bezier curves and has been applied to many digital images with good results.

TL;DR: Four measures for assessing the suitability of EOFs, quantifying their linearity, curvature bias, asymmetry, and overall goodness enable a better understanding to be gained of the individual merits of the EOF functions.

TL;DR: It is shown that both offset types of rational ruled surfaces are rational, and simple tool paths which are rational quartics are described.

TL;DR: This work introduces a Gaussian pyramid wavelet transform for multiresolution analysis of images, and shows that with a slight modification of the original Gaussian Pyramid a decomposition algorithm is yielded, which hasGaussian pyramid reduce and expand operations with a perfect reconstruction property.

TL;DR: This paper describes one approach to create a three dimensional object model with physically correct reflectance properties by observing a real object, and has successfully tested the approach using images of areal object.

TL;DR: This paper provides several postprocessing techniques for determining additional corner attributes (i.e., color, orientation, subtended angle, and contrast), which can be used to resolve otherwise ambiguous correspondences and to eliminate corners whose attributes do not match certain criteria.

TL;DR: Wavelet methods for the multiresolution representation of parametric curves and surfaces are developed and a new family of compactly supported symmetric biorthogonal wavelets with interpolating scaling functions are constructed.

TL;DR: It is proved that digital plane segments and their least-Squares plane fit are in one-to-one correspondence, which gives a simple representation of a digital plane segment by its base description and coefficients of the least-squares plane Fit, leading to a constant space representation of digital rectangles in space.

TL;DR: SpotIt! provides a virtually unlimited set of alternative features that can be browsed efficiently in the appropriate context, interactive holistic feature modification coupled to syntactic access to a feature database, and quantitative, automatic computation of face similarities.

TL;DR: For estimation of fractal dimension, the fractal Brownian function approach is extended to accommodate irregularly sampled data, and methods for segmenting sets of points exhibiting self-similarity over only certain scales are developed.

TL;DR: This paper describes how to construct a pyramid which has the property that if its vertex is translated to any point on a given rational Bezier surface patch with positive weights, the patch will lie completely outside the pyramid.

TL;DR: This paper interprets the through-the-lens camera control problem as a target tracking problem and formulate it as a constrianed nonlinear inversion problem that is then solved by integrating a tangent vector field defined on the configuration space of the virtual camera.

TL;DR: A zero-crossing-based 2D edge detector, which is the Laplacian of a rotationally invariant finite support function with a smooth boundary and yields optimal performance with respect to a penalty function defined in terms of a simplified SNR measure and Canny's edge localization, and spurious response measures.

TL;DR: This article proposes the use of computer-generated, yetgridless halftones, which involves solving the halftoning problem on the continuous 2D plane rather than on the usual discrete grid of pixels.

TL;DR: This paper presents a simple deterministic model based on a univariate iterative scheme which appears to emulate the behavior of auto models and allows it to be claimed that this instability coincides with structural changes in the corresponding spatial model and that the critical boundaries for the iteratives coincide with those for the spatial model.

TL;DR: A general scheme for noisy multi-edge detection is proposed and it is shown that the optimal smoothing filter in this case is the cascade of two infinite size symmetric exponential filters (ISEF), and an explicit analytical solution is given.

TL;DR: Subdivision methods for rectangular Bezier surfaces are generalized to subdivide a rectangular bezier surface patch of degreen×minto two rectangular BeZier surface patches of degreens× (m+n), while the parameter domain of the Beziers surface is decomposed into two trapezoids.

TL;DR: The fractal dimension of signals produced by sampled fBm, displaced interpolation, spectral synthesis, and spectral synthesis can be estimated with reasonable accuracy and analysis routines which make assumptions not made by the synthesis routine generally yield poor estimates of H, whereas a match of assumptions generally produces good estimates.