Analysis of Kannada Handwritten Vowels

Language identification for handwritten document images is an open document analysis problem. Handwritten character recognition has received extensive attention in academic and production fields. The recognition system can either online or offline. There is a large demand for handwritten character recognition and handwritten documents. India is a multi-lingual and multi-script country, where eighteen official scripts are accepted and have over hundred regional languages. Handwritten character recognition for Indian language is an important problem where there is relatively little work has been done. Particularly difficult is the problem of recognition of kagunita- the compound characters resulting from the consonant and the vowel combination. To recognize a kagunita, we need to identify the vowel and the consonant present in the kagunita character image. Unlike the Latin script used for the English language, it does not have upper case or lowercase. It has only one case of writing. Moreover, each alphabet contains more curves than straight lines. Hence handwritten Kannada character recognition is a challenging task. We had taken hundred handwritten datasets of different users. Handwritten Kannada vowels are scan and converted into binary image and normalized into a size of 64 x 64 pixels. Scanned image is segmented, we get a extracted of single Kannada vowels and then stored in a database. By using median filter we removed noise from the image. Using morphological thinning function, we get thinned Kannada vowels of different handwritten data sets. Overlapping the similar characters we get a overlapped image. By using morphological erosion and dilation function we get a standard representative image. Using this representative image, we can test different samples to get recognition accuracy. The performance of the proposed method is experimentally evaluated and the promising results and findings are presented.