Unconstrained Handwritten Numerals Research Articles

Research on Unconstrained Handwritten Numeral Recognition by BP Feature Screening Based on Fuzzy Clustering

In this paper, the recognition system of fuzzy clustering based on BP feature screening was put out. The figure specimens of experiment were filtered through BP network, and the result of screening was fit into the clustering source. At last fuzzy clustering was carried out by constituting the fuzzy relation matrix. The result of experiment demonstrates that this method has very high noise immunity capacity and overcame the limitation of traditional algorithm with single factor recognition. The recognition rate and precision ratio were greatly improved at the same time.

Off-line Recognition of Unconstrained Handwritten Numerals Using Fuzzy Hough Transform

A fuzzy system for handwritten numerals recognition using a fuzzy Hough transform technique is presented. The system is an off-line system since the data processed was written before the time of recognition. A data base of 480 patterns of unconstrained (free) handwritten numerals was used in the proposed system. Membership values are determined as fuzzy sets which are defined on the standard Hough transform vector. Manhattan distance measurement has been used to measure the similarity of an input feature vector to a number of numeral pattern classes. The overall recognition accuracy of the system for the ten numerals is 95%.

COMBINATION OF MULTIPLE CLASSIFIERS BY MINIMIZING THE UPPER BOUND OF BAYES ERROR RATE FOR UNCONSTRAINED HANDWRITTEN NUMERAL RECOGNITION

In order to raise a class discrimination power by the combination of multiple classifiers, the upper bound of Bayes error rate which is bounded by the conditional entropy of a class and decisions should be minimized. Based on the minimization of the upper bound of the Bayes error rate, Wang and Wong proposed only a tree dependence approximation scheme of a high-dimensional probability distribution composed of a class and patterns. This paper extends such a tree dependence approximation scheme to higher order dependency for improving the classification performance and thus optimally approximates the high-dimensional probability distribution with a product of low-dimensional distributions. And then, a new combination method by the proposed approximation scheme is presented and evaluated with classifiers recognizing unconstrained handwritten numerals.

Multiple-agents to identify and separate touching digits in unconstrained handwritten Hindi numerals

This paper presents a multiagent-based approach to the identification and recognition of touching in free handwritten Hindi numerals. We do not restrict our domain to touching pairs; rather, we consider numerals with arbitrary number of digits. Two agents are presented in this paper. The first agent works directly on the scanned image of the original handwritten number. It locates possible touching based on the thickness of handwriting. The other agent works on the thinned image. It segments the image into four categories of segments and tries to locate possible touching based on the rules that govern the connection of segments to form digits. After each of the two agents applies its own rules and investigates candidate possible touching cases, and to increase touching recognition rate, the two agents negotiate and try to agree on the actual touching cases. The experiments conducted so far are promising and successful. The obtained results are very encouraging with a success factor of 92.7%.

Multiple Classifier Fusion in Probabilistic Neural Networks

The main motivation of this paper is to design a statistically well justified and biologically compatible neural network model and, in particular, to suggest a theoretical interpretation of the well known high parallelism of biological neural networks. We consider a novel probabilistic approach to neural networks developed in the framework of statistical pattern recognition, and refer to a series of theoretical results published earlier. It is shown that the proposed parallel fusion of probabilistic neural networks produces biologically plausible structures and improves the resulting recognition performance. The complete design methodology based on the EM algorithm has been applied to recognise unconstrained handwritten numerals from the database of Concordia University Montreal. We achieved a recognition accuracy of about 95%, which is comparable with other published results.

Pattern recognition using Markov random field models

In this paper, we propose Markov random field models for pattern recognition, which provide a flexible and natural framework for modelling the interactions between spatially related random variables in their neighbourhood systems. The proposed approach is superior to conventional approaches in many aspects. This paper introduces the concept of states into Markov random filed models, presents a theoretic analysis of the approach, discusses issues of designing neighbourhood system and cliques, and analyses properties of the models. We have applied our method to the recognition of unconstrained handwritten numerals. The experimental results show that the proposed approach can achieve high performance.

MARKOV PROCESS IN PATTERN RECOGNITION

Using the Markov random process, we developed two new approaches to pattern recognition: (1) Hidden Markov model for modeling spectral features for recognizing 2D shapes. This is because Fourier spectra are suitable for describing 2D shapes of simple closed contours and probabilistic models are capable of coping with random variations in object shapes. We will analyze the properties of spectral features derived from contours of 2D shapes and use these features in 2D pattern recognition. (2) Markov random fields for modeling 2D structural and statistical features. We will give a theoretic analysis of this approach, discuss the issues in the design of neighborhood system and cliques for Markov random field models, and analyze the properties of the models. We have applied the proposed approach to the recognition of unconstrained handwritten numerals and 2D shapes. Our extensive experimental results show that the proposed approach can achieve a higher performance than that reported recently in the literature.

Hidden Markov models with spectral features for 2D shape recognition

We present a technique using Markov models with spectral features for recognizing 2D shapes. We analyze the properties of Fourier spectral features derived from closed contours of 2D shapes and use these features for 2D pattern recognition. We develop algorithms for reestimating parameters of hidden Markov models. To demonstrate the effectiveness of our models, we have tested our methods on two image databases: hand-tools and unconstrained handwritten numerals. We are able to achieve high recognition rates of 99.4 percent and 96.7 percent without rejection on these two sets of image data, respectively.

Application of Choquet Integral to Efficient Aggregation of Neural Networks

This paper presents a novel application of fuzzy integral to combine multiple neural networks subjectively. The concept of combining multiple neural networks or classifiers has been largely exploited in the fields of artificial neural networks and pattern recognition. The proposed method based on Choquet integral nonlinearly combines the outputs of the networks with subjective evaluation of the reliability of the individual neural networks. Experimental results with the recognition problem of totally unconstrained handwritten numerals show that the performance of the proposed method outperforms that of the conventional methods.

Pattern recognition with neural networks combined by genetic algorithm

Soft computing techniques have been recently exploited as a promising tool for achieving high performance in pattern recognition. This paper presents a hybrid method which combines neural network classifiers by genetic algorithm. Genetic algorithm gives us an effective vehicle to determine the optimal weight parameters that are multiplied by the network outputs as coefficients. The experimental results with the recognition problem of totally unconstrained handwritten numerals show that the genetic algorithm produces better results than the conventional methods such as averaging and Borda count.

Analysis of class separation and combination of class-dependent features for handwriting recognition

In this paper, we propose a new approach to combine multiple features in handwriting recognition based on two ideas: feature selection-based combination and class dependent features. A nonparametric method is used for feature evaluation, and the first part of this paper is devoted to the evaluation of features in terms of their class separation and recognition capabilities. In the second part, multiple feature vectors are combined to produce a new feature vector. Based on the fact that a feature has different discriminating powers for different classes, a new scheme of selecting and combining class-dependent features is proposed. In this scheme, a class is considered to have its own optimal feature vector for discriminating itself from the other classes. Using an architecture of modular neural networks as the classifier, a series of experiments were conducted on unconstrained handwritten numerals. The results indicate that the selected features are effective in separating pattern classes and the new feature vector derived from a combination of two types of such features further improves the recognition rate.

A three-dimensional neural network model for unconstrained handwritten numeral recognition: a new approach

The paper describes a three-dimensional (3-D) neural network recognition system for conflict resolution in recognition of unconstrained handwritten numerals. This neural network classifier is a combination of modified self-organizing map (MSOM) and learning vector quantization (LVQ). The 3-D neural network recognition system has many layers of such neural network classifiers and the number of layers forms the third dimension. The Experiments are conducted employing SOM, MSOM, SOM and LVQ, and MSOM and LVQ networks. These experiments on a database of unconstrained handwritten samples show that the combination of MSOM and LVQ performs better than other networks in terms of classification, recognition and training time. The 3-D neural network eliminates the substitution error.

A connectionist expert system model for conflict resolution in unconstrained handwritten numeral recognition

The paper describes a neural network and expert system model for conflict resolution of unconstrained handwritten numerals. The basic recognizer is the neural network. It solves most of the cases, but will fails in certain confusing cases. The expert system, the second recognizer, resolves the confusions identified by the neural network. The neural network classifier is a combination of Modified Self-Organizing Map (MSOM) and Learning Vector Quantization (LVQ). The experiments are conducted on Self-Organizing Map (SOM), MSOM, and the combinations of SOM & LVQ and MSOM & LVQ techniques. These experiments on a database of unconstrained handwritten samples show that the combination of MSOM & LVQ achieves satisfactory results in terms of classification, recognition and training time. The results obtained from this two-tier architecture are compared with the comments collected from an experiment conducted with a group of human experts specialized in unconstrained handwritten character recognition. The developed system is found to be useful in resolving conflicts in the recognition of Unconstrained Handwritten Numerals of PIN or ZIP codes of mailing addresses.

Recognition of unconstrained handwritten numerals by a radial basis function neural network classifier

Among the neural network models the RBF (Radial Basis Function) network seems to be quite effective for a pattern recognition task such as handwritten numeral recognition since it is extremely flexible to accommodate various and minute variations in data. Recently we obtained a good recognition rate for handwritten numerals by using an RBF network. In this paper we show how to design an RBF network classifier for a given problem in a well defined and easy-to-follow manner. We also report on the experiments to evaluate the performance of the RBF network classifier so designed.

Neural-network classifiers for recognizing totally unconstrained handwritten numerals

Artificial neural networks have been recognized as a powerful tool for pattern classification problems, but a number of researchers have also suggested that straightforward neural-network approaches to pattern recognition are largely inadequate for difficult problems such as handwritten numeral recognition. In this paper, we present three sophisticated neural-network classifiers to solve complex pattern recognition problems: multiple multilayer perceptron (MLP) classifier, hidden Markov model (HMM)/MLP hybrid classifier, and structure-adaptive self-organizing map (SOM) classifier. In order to verify the superiority of the proposed classifiers, experiments were performed with the unconstrained handwritten numeral database of Concordia University, Montreal, Canada. The three methods have produced 97.35%, 96.55%, and 96.05% of the recognition rates, respectively, which are better than those of several previous methods reported in the literature on the same database.

Multiresolution recognition of unconstrained handwritten numerals with wavelet transform and multilayer cluster neural network

In this paper, we propose a new scheme for multiresolution recognition of unconstrained handwritten numerals using wavelet transform and a simple multilayer cluster neural network. The proposed scheme consists of two stages: a feature extraction stage for extracting multiresolution features with wavelet transform, and a classification stage for classifying unconstrained handwritten numerals with a simple multilayer cluster neural network. In order to verify the performance of the proposed scheme, experiments with unconstrained handwritten numeral database of Concordia University of Canada, Electro-Technical Laboratory of Japan, and Electronics and Telecommunications Research Institute of Korea were performed. The error rates were 3.20%, 0.83%, and 0.75%, respectively. These results showed that the proposed scheme is very robust in terms of various writing styles and sizes.

Off-line recognition of totally unconstrained handwritten numerals using multilayer cluster neural network

In this paper, we propose a new scheme for off-line recognition of totally unconstrained handwritten numerals using a simple multilayer cluster neural network trained with the backpropagation algorithm and show that the use of genetic algorithms avoids the problem of finding local minima in training the multilayer cluster neural network with gradient descent technique, and improves the recognition rates. In the proposed scheme, Kirsch masks are adopted for extracting feature vectors and a three-layer cluster neural network with five independent subnetworks is developed for classifying similar numerals efficiently. In order to verify the performance of the proposed multilayer cluster neural network, experiments with handwritten numeral database of Concordia University of Canada, that of Electro-Technical Laboratory of Japan, and that of Electronics and Telecommunications Research Institute of Korea were performed. For the case of determining the initial weights using a genetic algorithm, 97.10%, 99.12%, and 99.40% correct recognition rates were obtained, respectively.

A neural network based dedicated thinning method

This paper presents a new application of a neural network model in the area of dedicated (handwritten numerals) binary pattern thinning. The thinning method has been tested on 6000 randomly selected images of unconstrained handwritten numerals. The test results show that the method introduces minimal structural deformity and the skeletal shape of the underlying pattern is retained.

Multilayer cluster neural network for totally unconstrained handwritten numeral recognition

In this paper, we propose a simple multilayer cluster neural network with five independent subnetworks for off-line recognition of totally unconstrained handwritten numerals. We also show that the use of genetic algorithms for avoiding the problem of finding local minima in training the multilayer cluster neural network with gradient descent technique reduces error rates.In order to verify the performance of the proposed multilayer cluster neural network, experiments with unconstrained handwritten numeral database of Concordia University of Canada, that of Electro-Technical Laboratory of Japan, and that of Electronics and Telecommunications Research Institute of Korea were performed. For the case of determining the initial weights of the multilayer cluster neural network randomly, the error rates were 2.90%, 1.50%, and 0.80%, respectively. And, for the case of determining the initial weights using a genetic algorithm, the error rates were 2.20%, 0.87%, and 0.60%, respectively.

A method of combining multiple experts for the recognition of unconstrained handwritten numerals

For pattern recognition, when a single classifier cannot provide a decision which is 100 percent correct, multiple classifiers should be able to achieve higher accuracy. This is because group decisions are generally better than any individual's. Based on this concept, a method called the Behavior-Knowledge Space Method was developed, which can aggregate the decisions obtained from individual classifiers and derive the best final decisions from the statistical point of view. Experiments on 46451 samples of unconstrained handwritten numerals have shown that this method achieves very promising performances and outperforms voting, Bayesian, and Dempster-Shafer approaches. >