Global and Local Skew Detection of Handwritten Gurmukhi Script

Abstract

ASTRACT This paper presents a new and efficient method for detection and removal of Global and Varying (Local) Skew. Varying (Local) Skew comes into existence in handwritten script. The proposed method is based on Hough transform and windowing. The proposed method divides the Skewed document into horizontal windows. Firstly, for each window the values of skew angle is calculated using Hough transform and then average of skew angles is estimated which represents the Skewness of document. Orientation of text is also calculated. Finally, Document is rotated based on average value of skew angle and orientation. The proposed method calculates skew for both printed and handwritten script. The proposed method can also calculate Multiskew. Experiments and results show that proposed method has high accuracy and can work with degraded papers