An approach to the language discrimination in different scripts using adjacent local binary pattern. Issue 5 (3rd September 2017)

• Record Type: Journal Article
• Authors: Brodić, D.; Amelio, A.; Milivojević, Z. N.
• Is Part Of: Journal of experimental & theoretical artificial intelligence. Volume 29:Issue 5(2017)
• Page Start: 929
• Page End: 947
• Publication Date: 2017-09-03
• Subjects: Adjacent local binary pattern -- cryptanalysis -- language discrimination -- optical character recognition -- statistical analysis; Artificial intelligence -- Periodicals; 006.3
• DOI: 10.1080/0952813X.2016.1264090
• Languages: English
• ISSNs: 0952-813X

Abstract

Abstract : The paper proposes a language discrimination method of documents. First, each letter is encoded with the certain script type according to its status in baseline area. Such a cipher text is subjected to a feature extraction process. Accordingly, the local binary pattern as well as its expanded version called adjacent local binary pattern are extracted. Because of the difference in the language characteristics, the above analysis shows significant diversity. This type of diversity is a key aspect in the decision-making differentiation of the languages. Proposed method is tested on an example of documents. The experiments give encouraging results.