An Adaptive Modular Neural Network With Application to Unconstrained Character Recognition

Mui, Lik

doi:10.1142/9789812797797_0011

Cited by 3 publications

(4 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Mui et.al. [9] presented two important characteristics in NN: learning and generalization. Learning process associates with network architecture that will change the connection structure (13) between units and signal strength in the connection structure.…”

Section: Image Classificationsmentioning

confidence: 99%

License Plate Recognition using Multi-cluster and Multilayer Neural Networks

Abdullah

Khalid

Yusof

et al. 2006

2006 2nd International Conference on Information &Amp; Communication Technologies

View full text Add to dashboard Cite

Vehicle license plat recognition has been a much studied research area in many countries. Due to the different types of license plates being used, the requirement of an automatic license plate recognition systemr is rather different for each country. In this paper, an automatic license plate recognition system is proposed for Malaysian vehicles with standard license plates based on image processing, feature extraction and neural networks. The image-processing library is developed in-house which we referred to as Vision System Development Platform (VSDP). Multi-Cluster approach is applied to locate the license plate at the right position while Kirsch Edge feature extraction technique is used to extract features from the license plates characters which are then used as inputs to the neural network classifier. The neural network model is the standard multilayered perceptron trained using the back-propagation algorithm. The prototyped system has an accuracy of more than 91%, however, suggestions to further improve the system are dtscussed in this paper based on the analysis of the error.

show abstract

Section: Image Classificationsmentioning

confidence: 99%

License Plate Recognition using Multi-cluster and Multilayer Neural Networks

Abdullah

Khalid

Yusof

et al. 2006

2006 2nd International Conference on Information &Amp; Communication Technologies

View full text Add to dashboard Cite

show abstract

“…Neural networks are inherently parallel architectures, but can also properly be combined by exploiting their learning capabilities for implementing a decision function combining the outputs of different classifiers [13]. A parallel combination of classifiers entirely based on neural networks is described in [15], where a modular neural network is built for the recognition of unconstrained handwritten digits. The network consists of three subsequent stages:…”

Section: Connectionist-based Classifiersmentioning

confidence: 99%

“…For instance, in [20], the q q q q q q q q q q q q q q q q q q q q q q q q q q q q q Output Error q q q q q q q q q q q q q q q q q q q q q q q q q q q q q Output Error q q q q q q q q q q q q q q q q q q q q q q q q q q q q q Output Error q q q q q q q q q q q q q q recognition of the substructure shape in Kanji characters is carried out by first indexing structural features extracted from the character, and afterwards by verifying a hypothesized model using the corresponding neural network. A related approach is proposed in [15], where the verification modules are based on MLP classifiers. The autoassociator-based architecture is also based on a set of MLP for verification, but the neural networks are trained so as to force the output to follow the input value.…”

Section: Autoassociator-based Classifiermentioning

confidence: 99%

A serial combination of connectionist-based classifiers for OCR

Francesconi

Gori

Marinai

et al. 2001

IJDAR

View full text Add to dashboard Cite

In this paper we describe the connectionistbased classification engine of an OCR system. The classification engine is based on a new modular connectionist architecture, where a multilayer perceptron (MLP) acting as a classifier is properly combined with a set of autoassociators -one for each class -trained to copy the input to the output layer. The MLP-based classifier selects a small group of classes with high score, that are afterwards verified by the corresponding autoassociators. The learning samples used to train the classifiers are constructed by means of a synthetic noise generator starting from few grey level characters labeled by the user.We report experimental results for comparing three neural architectures: an MLP-based classifier, an autoassociator-based classifier, and the proposed combined architecture. The experiments show that the proposed architecture exhibits the best performance, without increasing significantly the computational burden.

show abstract

“…Mui et al [7] presented two important characteristics in NN: learning and generalization. Learning process associates with network architecture that will change the connection structure between units and signal strength in the connection structure.…”

Section: Nn Mathematical Model and Algorithmsmentioning

confidence: 99%

Recognizing vehicle lubricant oil quality via neural network

Abdullah

Omar

Abdullah³

et al.

Proceedings of the Eighth International Symposium on Signal Processing and Its Applications, 2005.

View full text Add to dashboard Cite

Currently, measuring either vehicle's mileage or duration or either one does maintain lubricant viscosity. However, these judgments are inaccurate because there are many other factors like conductivity, humidity and viscosity that may affect the oil quality. This paper proposed one theory of monitoring viscosity quality with Neural Network (NN) modelling by introducing factors like temperature, shear stress and pressure. One deterministic objective will be highlighted that is to develop NN modelling based on those three factors. This research also introduces normalization approach called Along Channel and logarithmic function due to various range of data input. NN modelling, an off-line system is explicitly designed with Backpropagation Algorithm and Multilayer Feedforward Network for learning process while its weight is calculated based on Nguyen Widrow number and Genetic Algorithm. There were 310 sample data, which divided into two sets; 149 data for training set and the rest for testing and vice versa. The application performance has achieved up to 85.91% result approaching real viscosity value.

show abstract

An Adaptive Modular Neural Network With Application to Unconstrained Character Recognition

Cited by 3 publications

References 9 publications

License Plate Recognition using Multi-cluster and Multilayer Neural Networks

License Plate Recognition using Multi-cluster and Multilayer Neural Networks

A serial combination of connectionist-based classifiers for OCR

Recognizing vehicle lubricant oil quality via neural network

Contact Info

Product

Resources

About