Showing papers in "Pattern Recognition in 1999"

TL;DR: Results indicate that the normalised entropy measure provides significantly improved behaviour over a range of imaged fields of view.

TL;DR: It is shown that normalized color rgb, saturation S and hue H, and the newly proposed color models c 1 c 2 c 3 and l 1 l 2 l 3 are all invariant to a change in viewing direction, object geometry and illumination.

TL;DR: Hierarchical non-linear image registration is classified into three main groups, where the registration at successive levels in the hierarchy increases in data complexity, warp complexity, or model complexity.

TL;DR: An unsupervised texture segmentation method is presented, which uses distributions of local binary patterns and pattern contrasts for measuring the similarity of adjacent image regions during the segmentation process.

TL;DR: Two novel characteristics, datum point invariance and line feature matching, are presented in palmprint verification, which are proposed to detect whether a couple of palmprints are from the same palm.

TL;DR: The theoretical and experimental analyses prove that the proposed method has the ability to classify many types of image objects, and is particularly suitable for classifying seemingly similar objects with subtle di⁄erences.

TL;DR: A classification of fuzzy distances with respect to the requirements needed for applications in image processing under imprecision is proposed and new distances based on mathematical morphology are proposed.

TL;DR: This work presents a topology preserving 3D skeletonization method which computes both surface and curve values which will probably become as valuable a tool for shape analysis in 3D, as it is in 2D.

TL;DR: An alternative plot which is more revealing than an ROC plot is presented and a comparative index is described which allows one to take advantage of anything that may be known about the relative severity of the two kinds of misclassification.

TL;DR: This work is motivated by the idea of obtaining various shape features computed directly from the VCC without going to Cartesian-coordinate representation, and presents some results using real shapes.

TL;DR: This paper presents two algorithmic approaches to the point matching problem, in an attempt to reduce its computational complexity, while still providing a guarantee of the quality of the final match.

TL;DR: A new approach to the problem of graph and subgraph isomorphism detection from an input graph to a database of model graphs is proposed, based on a preprocessing step in which the model graphs are used to create a decision tree.

TL;DR: A new approach is proposed for bridge cleaning based on ridge positions instead of classical methods based on directional maps for minutiae filtering, and two novel criteria and related algorithms are introduced for validating the endpoints and bifurcations.

TL;DR: It is demonstrated that the wavelet correlation features contain more information than the intensity or the energy features of each color plane separately.

TL;DR: A new heuristic method based on the combination of two functions to evaluate the quality of fuzzy c-partitions produced by fuzzy clustering algorithms, and its effectiveness is compared to some existing cluster-validity criterion.

TL;DR: A novel faster search scheme of gravity-center template matching compared with the traditional search method in an image for human face detection, which significantly saves the time consumed in rough detection of human faces in a mosaic image.

TL;DR: The result shows that the Fourier-wavelet descriptor is an efficient representation which can provide for reliable recognition of complex patterns such as roadsigns, keys, aircrafts, characters, etc.

TL;DR: An automatic fruit recognition system and a review of previous fruit detection work are reported, able to recognize spherical fruits in natural conditions facing difficult situations: shadows, bright areas, occlusions and overlapping fruits.

TL;DR: A direct application of current research is that for any pose, the method can be used to compress the image of the object in any possible illumination and give a new framework on how to address illumination effects in computer vision in general.

TL;DR: The proposed inspection system has been implemented and tested with various types of solder joints in SMDs and the experimental results have verified the validity of this scheme in terms of speed and recognition rate.

TL;DR: A method for the automatic localization of text embedded in complex images permits to detect the spatial position and the skew of the text lines which are present in the scene and to return a binary representation of each text line.

TL;DR: This paper proposes parameter estimation methods related to monogrid and hierarchical MRF models that are similar to the expectation–maximization (EM) algorithm and looks at the maximum likelihood estimate of the parameters given a tentative labeling obtained at the previous step.

TL;DR: Registration using the GCP/GA technique is found to be signiicantly faster and of comparable accuracy than two popular techniques in the computer vision and medical imaging literature.

TL;DR: This method is able to extract different decision trees of high accuracy and comprehensibility from the trained neural network to understand the working of the neural network besides performing classification.

TL;DR: A new technique for elastic deformation restriction to particular object shapes of any closed planar curve using localized multiscale contour parameterization based on the 1D dyadic wavelet transform is proposed.

TL;DR: The seminal ideas underlying the elastic matching technique are reviewed, the practical implications of an integral formulation of the approach are considered, and a more general Bayesian interpretation of the method is explored in order to address issues that are less naturally resolved within a continuum mechanical setting.

TL;DR: An explicit relationship between the size of the LoG mask and the probability of edge detection error introduced by its approximation is derived, providing a strong basis for its stable implementation and showing the advantages that such analysis can confer.

TL;DR: This algorithm is directly motivated by the recently popular probabilistic Hough transform (PHT), however, it has some specific advantages in comparison with other kinds of PHT methods.

TL;DR: The results have demonstrated that the proposed 2D fuzzy approach outperforms the 2D nonfuzzy approach and the one-dimensional (1D) fuzzy partition approach.

TL;DR: Experimental results show that combination of the feature-based classifiers increases verification accuracy, and benefits of using forged signatures for training purposes have been assessed.