A serial combination of connectionist-based classifiers for OCR

Abstract

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.

Scholarly Communication and Academic Presses

“…The statistics of contacts to RM are encouraging. In the first year of publication (may 2000-april 2001) we have registered approximately 100,000 contacts, with a monthly average at the eve of 2001 of approximately 12,500 opened pages, which indicates a trend of approximately 150,000 contacts for the current year 12 . Readers are obviously mainly Italian, but with meaningful quotas of readers from the German linguistic area, from France, Spain and Latin America.…”

“…Our experience has been addressed to developing neural networks-based OCR systems [12] and to the problem of multi-class document processing [13], [14], [15]. In particular we have proposed a knowledge-based architecture for processing documents of a multiclass domain using a classification-based architecture, and a method for the construction of the knowledge, used to understand documents, by means of a small learning set of the domain documents.…”

A Hybrid Model for Recognizing Handwritten Bangla Characters using Support Vector Machine

“…In each fold, 680 word images are used for training and 340 word images are used for testing.In [9] a survey on OCR systems is provided by the authors where a long history of OCR systems is discussed. The classifiers used to recognize characters have evolved over time from Neural Network [10,11] to Support Vector Machines (SVM) [12]. An online recognition method for hand sketched symbol is presented in [13].…”

Combining Model-Based and Discriminative Approaches in a Modular Two-Stage Classification System: Application to Isolated Handwritten Digit Recognition

“…Indeed, the idea of multiple classifiers combination to treat ambiguity is presented in [8], but the proposed system combine only different model-based classifiers and is only tested on 2D artificial data. On the other hand, the combination of model-based and discriminative approaches is proposed in [5][8] [21][22] but their motivations are different. In [5], the model-based approach is used in a second stage to slightly improve the rejection capability of the MLP used at the first stage.…”

“…On the other hand, the combination of model-based and discriminative approaches is proposed in [5][8] [21][22] but their motivations are different. In [5], the model-based approach is used in a second stage to slightly improve the rejection capability of the MLP used at the first stage. In [21], the authors use only a few MLPs to improve the accuracy of the first classifier, which used a reduced number of prototypes.…”