Word-based handwritten Arabic scripts recognition using DCT features and neural network classifier

AlKhateeb, Jawad H.; Ren, Jinchang; Jiang, Jianmin; Ipson, Stanley S.; Abed, Haikal El

doi:10.1109/ssd.2008.4632863

Cited by 35 publications

(30 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recognition rate was 91.70% on the IFN/ENIT Arabic standard database. AlKhateeb 2011 [6] used DCT features which extracted from each word sample, then features are fed to train a neural network for classification. The proposed system tested on IFN/ENIT Database each time 80% of the samples in the database are used for training and the remaining 20% for testing.…”

Section: Related Workmentioning

confidence: 99%

Proposed Handwriting Arabic Words classification Based On Discrete Wavelet Transform and Support Vector Machine

2017

ijs

View full text Add to dashboard Cite

A proposed feature extraction algorithm for handwriting Arabic words. The proposed method uses a 4 levels discrete wavelet transform (DWT) on binary image. sliding window on wavelet space and computes the stander derivation for each window. The extracted features were classified with multiple Support Vector Machine (SVM) classifiers. The proposed method simulated with a proposed data set from different writers. 1.IntroductionThe handwriting recognition process means converting the handwriting text images into understandable text by the computer and by many applications such as postal address reading for mail sorting purposes, cheques recognition and word spotting on a handwritten text page [1].Offline recognition to Arabic handwriting is still very challenging because the writing styles is defferted from person to other, Also for the same person at different times . Because of the coursive nature of the Arabic script and the similar appearance of some Arabic characters makes the handwriting recognition difficult task. the recognition system has several stages such as Image acquisition, preprocessing, feature extraction and classification / recognition.Achieving high recognition accuracy depends mainly on the used feature selection method [2].The first stage in any recognition system is preprocessing which tries to reduce the noise data and keep only the desired information and make the next operation (feature extraction process) easy to implement. Moreover, the second stage is feature extraction and selection; they are extracting useful information from the binary handwriting word

show abstract

Section: Related Workmentioning

confidence: 99%

Proposed Handwriting Arabic Words classification Based On Discrete Wavelet Transform and Support Vector Machine

2017

ijs

View full text Add to dashboard Cite

show abstract

“…Where and One of the characteristics of DCT is its ability to convert the energy of the image into a few coefficients [14]. DCT is applied on the low-frequency sub-band (LL) of DWT image to extract DCT coefficients.…”

Section: Discrete Cosine Transformmentioning

confidence: 99%

Handwritten Digit Recognition based on DWT and DCT

Lawgali¹

2015

IJDTA

View full text Add to dashboard Cite

show abstract

“…This is a relatively high recognition rate, though the number of classes that needed to be recognised (34) was relatively low. In a similar study, whole word recognition was applied to 500 Tunisian town/village names using an artificial neural network and a recognition accuracy of 82.5% was achieved [3]. In another study involving Arabic character recognition, in addition to moment features, the "number of dots" and "number of holes" were also extracted, which described the diacritics in more detail [1].…”

Section: Related Workmentioning

confidence: 99%

“…The 2-D coefficients of the DCT are converted to a 1-D vector using a technique called zig-zagging [3] (Figure 7 (b)), the purpose of which is to extract the low frequency coefficients from the DCT first. In this study, the extraction of features varied in terms of the size of the cells that the image or sliding window was partitioned into for the DCT-II calculation and the number of DCT coefficients used as features.…”

Section: Discrete Cosine Transformmentioning

confidence: 99%

A comparison of machine learning techniques for handwritten |Xam word recognition

Williams

Suleman

Paihama

2013

Proceedings of the South African Institute for Computer Scientists and Information Technologists Conference

View full text Add to dashboard Cite

The Bleek and Lloyd collection contains 19th century handwritten notebooks that document the language and culture of the |Xam-speaking people who lived in Southern Africa. Access to this rich data could be enhanced by transcriptions of the text; however, the complex diacritics used in the notebooks complicate the process of transcription. Machine learning techniques could be used to perform this transcription, but it is not known which techniques would produce the best results. This paper thus reports on a comparison of 3 popular techniques applied to this problem: artificial neural networks (ANN); hidden Markov models (HMM); and support vector machines (SVM). It was found that an SVM-based classifier using histograms of oriented gradients as features resulted in the best word recognition accuracy of 58.4%. Furthermore, it was found that most feature extraction parameters did not have a large effect on recognition accuracy and that the SVM-based recognisers outperform both ANN-and HMM-based recognisers.

show abstract

Word-based handwritten Arabic scripts recognition using DCT features and neural network classifier

Cited by 35 publications

References 15 publications

Proposed Handwriting Arabic Words classification Based On Discrete Wavelet Transform and Support Vector Machine

Proposed Handwriting Arabic Words classification Based On Discrete Wavelet Transform and Support Vector Machine

Handwritten Digit Recognition based on DWT and DCT

A comparison of machine learning techniques for handwritten |Xam word recognition

Contact Info

Product

Resources

About