Feature extraction methods for character recognition-A survey

Trier, Øivind Due; Jain, Anil K.; Taxt, Torfinn

doi:10.1016/0031-3203(95)00118-2

Cited by 973 publications

(388 citation statements)

References 60 publications

Supporting

Mentioning

377

Contrasting

Unclassified

Order By: Relevance

“…We perform slant/skew/baseline normalizations that are commonly used in the literature (for example, see [11]). While some of our features are generally used for the recognition of handwritten characters [17], we use them to represent the shape of entire words.…”

Section: Related Workmentioning

confidence: 99%

Holistic word recognition for handwritten historical documents

Lavrenko

Rath

Manmatha

First International Workshop on Document Image Analysis for Libraries, 2004. Proceedings.

168

143

View full text Add to dashboard Cite

show abstract

Section: Related Workmentioning

confidence: 99%

Holistic word recognition for handwritten historical documents

Lavrenko

Rath

Manmatha

First International Workshop on Document Image Analysis for Libraries, 2004. Proceedings.

168

143

View full text Add to dashboard Cite

show abstract

“…The features vector size is also important in order to avoid a phenomenon called the dimensionality problem. Several methods for features extraction are designed for different representations of the characters, such as binary characters, character contour, skeletons (thinned characters), or even gray levels characters [16]. The features extraction methods are valued in terms of invariance properties, and expected distortions and variability of the character.…”

Section: Optical Characters Recognition Systemsmentioning

confidence: 99%

A Set of Features Extraction Methods for the Recognition of the Isolated Handwritten Digits

Ouchtati¹,

Redjimi²,

Bedda³

2014

IJCCE

View full text Add to dashboard Cite

Abstract-In this paper we present an off line system for the recognition of the isolated handwritten digits. The study is based mainly on the evaluation of neural network performances, trained with the gradient back propagation algorithm and fed by several feature vectors. The used parameters to form the input vector of the neural network are extracted on the binary images of the digits by several methods: the distribution sequence, sondes application, the Barr features, The central moments of image coding according to the directions of Freeman, and the centred moments of the different projections and profiles.

show abstract

“…• Linear kernel (8) • Polynomial kernel (9) • RBF (Gaussian) kernel (10) • Sigmoidal kernel (11) where are data vectors in input space. In the current experiment, RBF kernel function is proposed as a choice for identifying RNA samples, and it was found to give good classification performance.…”

Section: ) Binary Classification By a Svmmentioning

confidence: 99%

“…The ultimate aim of our work is to perform automatic classification of data obtained from T-ray measurements with tomographic applications [7]. It is important to devise effective feature extraction methods to fully represent the different characteristics of these signals [8]. Signal processing methods are proposed for the current experiment.…”

Section: Introductionmentioning

confidence: 99%