Handwritten Bangla character recognition using a soft computing paradigm embedded in two pass approach

作者：

Highlights：

• A soft computing paradigm is designed within the framework of Two pass approach.

• The technique is used to evaluate the recognition of handwritten Bangla characters.

• The handwritten Bangla characters consist of Basic and Compound and Allographs.

• An algorithmic methodology is developed for formation of pattern groups.

• GA based local region selection method is used to improve recognition accuracies.

摘要

•A soft computing paradigm is designed within the framework of Two pass approach.•The technique is used to evaluate the recognition of handwritten Bangla characters.•The handwritten Bangla characters consist of Basic and Compound and Allographs.•An algorithmic methodology is developed for formation of pattern groups.•GA based local region selection method is used to improve recognition accuracies.

论文关键词：Handwritten Bangla characters recognition,Bangla Compound characters,Genetic algorithm,Soft computing paradigm embedded in Two pass approach

论文评审过程：Received 14 November 2013, Revised 20 November 2014, Accepted 10 December 2014, Available online 31 December 2014.

论文官网地址：https://doi.org/10.1016/j.patcog.2014.12.011