查看文章

In this paper, a problem of automatic selection of feature sets and suitable classifiers for recognition of handwritten scripts at word level is addressed. The problem is brought out clearly with sufficient study by comparing state of the art techniques. Based on the analysis, three different models have been proposed in this paper. To accomplish the task, combination of various features and classifiers are tried out. The proposed work is on bi-script recognition. To conduct experimentation we have considered different features and classifiers which are recommended in the literature. The proposed work has been demonstrated for its effectiveness on our own dataset of reasonably large size. Experimental results reveal that, the proposed models perform better than the existing models.