Pattern Recognition Theory Research Articles

GAMEs: Growing and adaptive meshes for fully automatic shape modeling and analysis

This paper presents a new framework for shape modeling and analysis, rooted in the pattern recognition theory and based on artificial neural networks. Growing and adaptive meshes (GAMEs) are introduced: GAMEs combine the self-organizing networks which grow when require (SONGWR) algorithm and the Kohonen's self-organizing maps (SOMs) in order to build a mesh representation of a given shape and adapt it to instances of similar shapes. The modeling of a surface is seen as an unsupervised clustering problem, and tackled by using SONGWR (topology-learning phase). The point correspondence between point distribution models is granted by adapting the original model to other instances: the adaptation is seen as a classification task and performed accordingly to SOMs (topology-preserving phase). We thoroughly evaluated our method on challenging synthetic datasets, with different levels of noise and shape variations. Finally, we describe its application to the analysis of a challenging medical dataset. Our method proved to be reproducible, robust to noise, and capable of capturing real variations within and between groups of shapes.

Automatic signal detection based on support vector machine

Algorithm of STA/LTA is frequently used in automatic signal detection, in which the range of detection threshold is (0, ∞), the optimal threshold should be determined by experiment to make a balance between false detection and missing detection. By using the theory of pattern recognition, a new algorithm for automatic signal detection based on support vector machine was proposed and the method of preprocess and pattern feature extraction were discussed as well as the selection of kernel function for support vector machine. The detection performance of the new algorithm was analyzed by means of real seismic data. The experiments showed that the new method could simplify the selection of threshold and detect signal accurately. In addition to the better performance of anti-noise, the ratio of false detection could decrease 85% in comparison with that of STA/LTA.

Correlation pattern recognition

Correlation is a robust and general technique for pattern recognition and is used in many applications, such as automatic target recognition, biometric recognition and optical character recognition. The design, analysis and use of correlation pattern recognition algorithms requires background information, including linear systems theory, random variables and processes, matrix/vector methods, detection and estimation theory, digital signal processing and optical processing. This 2005 book provides a needed review of this diverse background material and develops the signal processing theory, the pattern recognition metrics, and the practical application know-how from basic premises. It shows both digital and optical implementations. It also contains technology presented by the team that developed it and includes case studies of significant interest, such as face and fingerprint recognition. Suitable for graduate students taking courses in pattern recognition theory, whilst reaching technical levels of interest to the professional practitioner.

Online fatigue damage monitoring by ultrasonic measurements: A symbolic dynamics approach

The paper presents an analytical tool for early detection and online monitoring of fatigue damage in polycrystalline alloys that are commonly used in mechanical structures of human-engineered complex systems. Real-time fatigue damage monitoring algorithms rely on time series analysis of ultrasonic signals that are sensitive to micro-structural changes occurring inside the material during the early stages of fatigue damage; the core concept of signal analysis is built upon the principles of Symbolic Dynamics, Statistical Pattern Recognition and Information Theory. The analytical tool of statistical pattern analysis has been experimentally validated on a special-purpose test apparatus that is equipped with ultrasonic flaw detection sensors and a travelling optical microscope. The paper reports fatigue damage monitoring of 7075-T6 alloy specimens, where the experiments have been conducted under load-controlled constant amplitude sinusoidal loadings for low-cycle and high-cycle fatigue.

Filter learning: Application to suppression of bony structures from chest radiographs

A novel framework for image filtering based on regression is presented. Regression is a supervised technique from pattern recognition theory in which a mapping from a number of input variables (features) to a continuous output variable is learned from a set of examples from which both input and output are known. We apply regression on a pixel level. A new, substantially different, image is estimated from an input image by computing a number of filtered input images (feature images) and mapping these to the desired output for every pixel in the image. The essential difference between conventional image filters and the proposed regression filter is that the latter filter is learned from training data. The total scheme consists of preprocessing, feature computation, feature extraction by a novel dimensionality reduction scheme designed specifically for regression, regression by k-nearest neighbor averaging, and (optionally) iterative application of the algorithm. The framework is applied to estimate the bone and soft-tissue components from standard frontal chest radiographs. As training material, radiographs with known soft-tissue and bone components, obtained by dual energy imaging, are used. The results show that good correlation with the true soft-tissue images can be obtained and that the scheme can be applied to images from a different source with good results. We show that bone structures are effectively enhanced and suppressed and that in most soft-tissue images local contrast of ribs decreases more than contrast between pulmonary nodules and their surrounding, making them relatively more pronounced.

Aspects of recent developments in analytical chemometrics

Some aspects of recent developments in analytical chemometrics are discussed, in particular the developments viewed from the angle of the research efforts undertaken in authors’ laboratories. The topics concerned include resolution of high-order chemical data, morphological theory and methodology for chemical signal processing, multivariate calibration and chemical pattern recognition for solving complex chemical problems, and resolution of two-way chemical data from hyphenated chromatographic instruments.

Problems of solvability and choice of algorithms for decision making by precedence

A wide range of practical and theoretical problems are reduced to decision making, which is interpreted rather ambiguously from the viewpoint of its statement. This is mainly related to the fact that the concept of decision cannot be formalized. From the methodological point of view, decision making is a process whose result is a decision. Even if the concept of decision is not defined, such a methodological view of the problem highlights its key points in a different way. In this case, the problem is transformed into a sequence of clearer and unambiguously defined problems. In this sequence, there certainly exists the problem of calculating the property of objects, and the decision itself can be associated with such a property [21]. Then, the problem of decision making can be reduced to calculating (determining) the properties of the information analyzed. This is undoubtedly one version of the problem of decision making, which is based on the problem of determining the above properties. Decision making allows one to easily find out if all property determination problems have a common nature and to reduce them to a single statement on a formal level. What problems are meant by the determination of properties? These are well-studied problems of sentential calculus and predicate calculus [1, 2]; problems of logical diagnostics [3, 4], which are less studied but have a larger number of practical interpretations; and, finally, just poorly formalized problems in pattern recognition theory (PRT) [5, 9]. It is obvious that, formally, all the above problems contain sets divided into subsets. Each subset is characterized by a certain property. The problem consists in calculating this property for each element of the original set.

Multi-objectives fuzzy optimization model for ship form demonstration based on information entropy

Selecting optimization ship form scheme is an important content in the process of concept design of ship. Multi-objective fuzzy decision-making model for ship form demonstration is set up according to the fuzzy pattern-recognition theory. Weight coefficients of each target of ship form scheme are determined by information entropy and individual subjective partiality. This model is used to select the optimal ship form scheme, the example shows that the model is exact and the result is credible. It can provide a reference for choosing the optimization scheme of ship form.

A Problem of Pattern Recognition in Arrays of Interrelated Objects. Recognition Algorithm

The classical methods of pattern recognition theory presuppose the independence of elements of a recognizable set. Therefore, even in the case of interrelated objects, the decision on the class of each object is taken independent of the decision on the classes of other objects. At the same time, it is obvious that in this case it is necessary to make consistent decisions on the classes of elements of the interrelated array. In particular, this necessity stems from the prior supposition that neighboring objects of the interrelated array more often belong to one class than to different classes. In many practical problems, such a prior supposition proves to be very natural, enabling one to develop effective recognition algorithms.

Pattern Recognition Theory and Preliminary Processing of Images Based on Stochastic Geometry

New geometrical dual trace transformation connected with scanning images with complex trajectories is introduced. On the basis of this transformation, non-linear filtration is accomplished in order to reduce the noise content of images, segmentation, smoothing, and polygonal approximation.

A new way of looking at eating

the analysis of meals, especially of spontaneous meals, has long had a problematic place in the behavioral neuroscience of eating. Only a small minority of eating researchers now, or ever, have concerned themselves with meal patterns. On first glance this seems paradoxical. After all, eating occurs

Moving targets pattern recognition based on the wavelet neural network

Based on pattern recognition theory and neural network technology, moving objects automatic detection and classification method integrating advanced wavelet analysis are discussed in detail. An algorithm of moving targets pattern recognition on the combination of inter-frame difference and wavelet neural network is presented. The experimental results indicate that the designed BP wavelet network using this algorithm can recognize and classify moving targets rapidly and effectively.

Self-Formation Supported by Pattern Recognition

Artificial planar systems self-formation process supported by pattern recognition theory and methods are discussed. Concept possibilities to apply pattern recognition power for improving and control self-formation processes are presented.

A similarity function to evaluate the orthodontic condition in patients with cleft lip and palate

The objective of this work is the modeling of a similarity function adapted to the medical environment using the logical-combinatorial approach of pattern recognition theory, and its application to compare the orthodontic conditions of patients with cleft-primary palate and/or cleft-secondary palate congenital malformations. The variables in domains with no a priori algebraic or topological structure are objects whose similarity or difference is evaluated by comparison criteria functions. The range of these functions is an ordered set normalized into the unit interval, and they are designed to allow differentiation and non-uniform treatment of the object-variables. The analogy between objects is formalized as a similarity function that stresses the relations among the comparison criteria and evaluates the partial descriptions (partial similarity/difference) or total descriptions (total similarity/difference) of the objects. For the orthodontic problem we defined a set of 12 variables featuring the unilateral/bilateral fissures, the conditions of maxilla, premaxilla, mandible and patient's bite. The comparison criteria (logical for malocclusion, fuzzy for maxillary collapse unilateral/anteroposterior and for overbite, and Boolean for protrusive/retrusive premaxilla conditions) were assigned a relevance factor based on the orthodontist accumulated knowledge and experience. The modeling of the similarity function and its effectiveness in comparing orthodontic conditions in patients are illustrated by the study of four clinical cases with different clefts. The results through similarity are close to the expected ones. Moreover evaluated at different moments it allows to assess the effect of treatment in a single patient, hence providing valuable auxiliary criteria for medical decision making as to the patient's rehabilitation. We include the potential extension of the methodology to other medical disciplines such as speech therapy and reconstructive surgery.

An application of pattern recognition on scoring Chinese corporations financial conditions based on backpropagation neural network

The paper follows three parts on the whole. The first part reviews the literature on financial diagnosis and the appropriate measures to be currently used. Thus the research gaps in the fields are highlighted. Next part details the variables and samples selection, data, analysis and results applying to scoring financial conditions of a Chinese corporation. By using pattern recognition theory, a scoring model is developed to analyze corporate financial conditions by backpropagation neural networks. It has been proved to be better than ordinary BPNN and traditional multivariate discriminant analysis. Finally, conclusions are presented.

A multi-dimensional fuzzy decision support strategy

Based on the theory of fuzzy pattern recognition and fuzzy optimization, this paper presents a kind of multi-dimensional fuzzy decision support strategy for multi-objective and multi-layer fuzzy decision support systems. Compared with conventional strategies of fuzzy optimization that just considers the relative membership u j of decision j of down layer unit systems, this strategy also takes the relative bad membership u j b=1− u j as the input of up layer unit systems. Besides, unlike conventional strategies of fuzzy optimization, this strategy can provide multi-dimensional information to the decision-maker. With these advantages, this strategy founds a new theoretical basis for the multi-dimensional decision making of large systems.

A Pattern Recognition Approach for Manufacturing Facility Compaction by Machining Function Combination Using Flexible Manufacturing Modules

Facility layout and flexible automation are two approaches for reduction of material handling costs and space requirements in a machining facility that have always been implemented independently of each other. This paper describes an integrated approach to compaction of existing machining facilities using machine grouping algorithms and multi-function machining centers, also referred to as Flexible Machining Modules (FMMs). First, we decompose a facility into a network of layout modules to reduce product travel distances and simplify material flow control. Then, subject to design feasibility, we identify those sets of machines in each module that could be replaced by multi-function FMMs that could be linked into a Flexible Machining System (FMS). The proposed approach uses a combination of pattern recognition and graph theory algorithms utilized for facility layout. The paper concludes with a description of a validation study conducted in an aerospace machining jobshop.

On learning multicategory classification with sample queries

Consider the pattern recognition problem of learning multicategory classification from a labeled sample, for instance, the problem of learning character recognition where a category corresponds to an alphanumeric letter. The classical theory of pattern recognition assumes labeled examples appear according to the unknown underlying pattern-class conditional probability distributions where the pattern classes are picked randomly according to their a priori probabilities. In this paper we pose the following question: Can the learning accuracy be improved if labeled examples are independently randomly drawn according to the underlying class conditional probability distributions but the pattern classes are chosen not necessarily according to their a priori probabilities? We answer this in the affirmative by showing that there exists a tuning of the sub-sample proportions which minimizes a loss criterion. The tuning is relative to the intrinsic complexity of the Bayes-classifier. As this complexity depends on the underlying probability distributions which are assumed to be unknown, we provide an algorithm which learns the proportions in an on-line manner utilizing sample querying which asymptotically minimizes the criterion. In practice, this algorithm may be used to boost the performance of existing learning classification algorithms by apportioning better sub-sample proportions.

SubIslands: the probabilistic match assignment algorithm for subcircuit recognition

The recognition of subcircuit instances in a larger circuit is widely used in the simulation, verification, and testing of integrated circuit computer-aided designs. Subcircuit recognition (SR) can be stated as a problem of finding images of a small model bipartite graph (BG) corresponding to a subcircuit in a large object BG corresponding to a circuit. The best-known SR algorithms are based on the search-oriented subgraph isomorphism methods. Unfortunately, these methods may require a long runtime for large and highly symmetrical circuits. The authors develop a new high-performance probabilistic recognition method for solving the SR problem. This method combines: 1) the graduated assignment matching technique; 2) two well-known concepts from pattern recognition theory, namely, the error propagation and the delayed decision making; and 3) an efficient probabilistic BG labeling algorithm. In contrast to the search-based algorithms, the new recognition method solves the SR problem as an optimization task and, as a consequence, allows extremely fast simultaneous finding of subcircuit images. The experimental results show that this approach recognizes all the subcircuit instances orders of magnitude faster than the search-oriented algorithms.

Blood glucose dynamics and control of meal initiation: a pattern detection and recognition theory.

A new framework for understanding the control of feeding behavior, with special emphasis on the evolution of hunger, the initiation of feeding, and its dependence on patterns of blood glucose, is the subject of this review. A perspective on the current status and future directions of this search for a more complete understanding of the regulation of feeding behavior in laboratory rats and humans is presented including theoretical and experimental components. First, a historical perspective on the role of blood glucose in the control of feeding is presented. Next, the theoretical approaches that have been applied to the control of feeding and had a strong influence on experimental feeding research are summarized. This is followed by a statement and overview of a current theory that has emerged from studies of the role of transient declines in blood glucose in the control of meal initiation. The current working hypothesis that transient declines in blood glucose are endogenous metabolic patterns that are detected and recognized by the central nervous system and are mapped into meal initiation in rats and are correlated with meal requests in humans are then presented. Then, the experimental studies on meal initiation and its dependence on patterns of blood glucose, first in rats and then in humans, are reviewed in detail. Finally, the future directions of the work, limitations, and the implications for the understanding of the control of feeding behavior and the regulation of energy balance are discussed.