International Journal of Science and Research (IJSR)

International Journal of Science and Research (IJSR)
Call for Papers | Fully Refereed | Open Access | Double Blind Peer Reviewed

ISSN: 2319-7064


Downloads: 121 | Views: 183

M.Tech / M.E / PhD Thesis | Computer Science & Engineering | India | Volume 4 Issue 7, July 2015


Mixture of Printed and Handwritten Kannada Numeral Recognition Using Normalized Chain Code and Wavelet Transform

Shashikala Parameshwarappa | B.V.Dhandra


Abstract: Optical character recognition (OCR) is one of the most successful applications of automatic pattern recognition. The current research in OCR is now addressing documents that are not well handled by the available systems, including severely degraded, omnifont machine-printed text and (unconstrained) handwritten text. In this paper, a novel method for recognition of printed and handwritten mixed isolated Kannada numeral is presented. An algorithm is proposed to recognize the printed and handwritten Kannada numerals based on shape features such as normalized chain codes and wavelet filters. A normalized chain code and two-dimensional discrete wavelet transforms are proposed to extract as a feature vector of size 22 from the normalized binary images of size 64x64. The SVM and KNN classifier with 2 fold cross validation is applied for classification of handwritten and printed mixed Numerals. The proposed algorithm is experimented on a data set of 4000 numeral images consisting of handwritten and printed numerals. Further the proposed system achieves the average recognition accuracy of 98.04 % and 99.12 % for mixed Numerals by KNN and SVM classifiers respectively. It achieves reasonably high recognition accuracy with less number of features set.


Keywords: Kannada Numerals, OCR, Normalized chain code, wavelet transform, SVM classifier


Edition: Volume 4 Issue 7, July 2015,


Pages: 1453 - 1457


How to Download this Article?

Type Your Valid Email Address below to Receive the Article PDF Link


Verification Code will appear in 2 Seconds ... Wait

Top