Isolated Handwritten Devnagri Character Recognition using Fourier Descriptor and HMM

This paper describes a complete system for the reco gnition of isolated handwritten Devnagri character using Fourier Descri ptor and Hidden-Markov model (HMM). The HMM has the property that its stat es are not defined as a priory information, but are determined automaticall y based on a database of handwritten numerals images. In this work the image database consist of 500 images of handwritten Devnagri characters from 50 dwriters. Before extracting the features, the images are normalized using image isometrics such as translation, rotation and scaling. After normalizat ion the Fourier features are extracted using Fourier Descriptor. An automatic sy stem trained 400 images of image database and character model form with multiv ariate Gaussian state conditional distribution. A separate set of 100 cha racters was used to test the system. The recognition accuracy for individual cha racter varies from 90% to 100% for number of states per model N=3 and 100% fo r N=5