摘要
Given the number and variety of methods used for handwriting recognition, ithas been shown that there is no single method that can be called the ''best''. In recent years, thecombination of different classifiers and the use of contextual information have become major areasof interest in improving recognition results. This paper addresses a case study on the combinationof multiple classifiers and the integration of syntactic level information for the recognition ofhandwritten Arabic literal amounts. To the best of our knowledge, this is the first time either ofthese methods has been applied to Arabic word recognition. Using three individual classifiers withhigh level global features, we performed word recognition experiments. A parallel combination methodwas tested for all possible configuration cases of the three chosen classifiers. A syntacticanalyzer makes a final decision on the candidate words generated by the best configuration scheme.The effectiveness of contextual knowledge integration in our application is Arabicconfirmed by theobtained results.
Given the number and variety of methods used for handwriting recognition, ithas been shown that there is no single method that can be called the ''best''. In recent years, thecombination of different classifiers and the use of contextual information have become major areasof interest in improving recognition results. This paper addresses a case study on the combinationof multiple classifiers and the integration of syntactic level information for the recognition ofhandwritten Arabic literal amounts. To the best of our knowledge, this is the first time either ofthese methods has been applied to Arabic word recognition. Using three individual classifiers withhigh level global features, we performed word recognition experiments. A parallel combination methodwas tested for all possible configuration cases of the three chosen classifiers. A syntacticanalyzer makes a final decision on the candidate words generated by the best configuration scheme.The effectiveness of contextual knowledge integration in our application is Arabicconfirmed by theobtained results.
