Handwritten Khmer text recognition
Bayram Annanurov, Norliza Mohd Noor · 2016
This paper proposes a model for an offline handwritten Khmer character recognition. We make use of two dimensional Fourier transformation for feature selection and feed-forward Artificial Neural Net as classification tool. The recognition system allows using the nature of Khmer writing, which is an example of alphasyllabary (Abugida) writing systems. The recognition of the normalized handwritten images has been performed for comparison purposes. The results suggest that the recognition rate increases with reduced feature set.