2026-10-17T11:34:53Z https://meral.edu.mm/oai

oai:meral.edu.mm:recid/4482 2021-12-13T03:29:23Z 1582963302567:1597824273898 user-ucsy

Handwritten Character Recognition based on Competitive Neural Trees Htike, Theingi Thein, Yadana Competitive Neural Trees (CNeT) are widely used forclassification in pattern recognition. This paperapplies this technique for recognizing of Myanmarhandwritten characters. This paper involves threeimportant steps, typically preprocessing, featureextraction and classification. The aim ofpreprocessing is to improve the quality of the imagesfor further processing. For the extraction of features,four of the 3×3 masks are applied to word images toextract horizontal, vertical, right and left-diagonallines. Afterwards, decomposed images should bepartitioned to eight sectors around the center ofimage and the number of black pixels in each sectorcalculated and normalized by dividing them upon thetotal number of black pixels in word images forfeature vector. These feature vector from word imageswhich are used in Competitive Neural Trees (CNeT)for recognition purpose. This paper introduces aglobal search method for the CNeT, which is utilizedfor training. 2012-02-28 http://hdl.handle.net/20.500.12678/0000004482 https://meral.edu.mm/records/4482