Slant Correction and Resampling for Online Handwritten Characters Recognition of Gurmukhi Script

Abstract

Handwritten characters of natural handwriting are usually italicized due to mechanism of handwriting and personality. In this thesis the slant of the Gurmukhi characters has been estimated.Later resampling has also been done. In slant correction and estimation, we have used chain code algorithm. We have shown the comparison of the results obtained by 8-directional and 16- direction chain code algorithm. Bezier interpolation method has been used in resampling of points in a stroke. The purpose of the thesis is to focus slant correction and resampling, which are two important steps in preprocessing phase of online handwriting recognition. Bezier Interpolation method has been used for Resampling.