Handwritten Numerals Recognition using Hough Transformation Technique

Munish Kumar, Deepika Ahuja

Abstract


Handwriting recognition has always been a exigent task in image processing and pattern recognition. In this paper, we have presented Hough transformation technique for extracting the features of handwritten numerals. SVM based approach has been used to classify the numerals based on the extracted features. In this study, we have taken the samples of handwritten numerals from 120 different writers. Each writer has written each numeral five times. We have collected total 6000 samples of handwritten numerals for training and testing. We have taken 75% data as training data and 25% data as testing data. The proposed system achieves a maximum recognition accuracy of about 91.75%.


Keywords - SVM classifier, Hough transform, OCR.


Full Text:

PDF


DOI: https://doi.org/10.26483/ijarcs.v3i5.1371

Refbacks

  • There are currently no refbacks.




Copyright (c) 2016 International Journal of Advanced Research in Computer Science