A Fast High Precision Algorithm for the Estimation of Skew Angle Using Moments

摘要

An efficient, high precision algorithm is proposed for the correction of the skew angle of scanned document images used in OCR (Optical Character Recognition). The algorithm employs moments to find the primary axis of every object in the document rather than the Hough transform and handles arbitrary angles. The object can be a halftone image or a line of text. Lines are described using the Freeman's chain code. A simple technique used to find the primary axis of each object from the chain code. Using a feature that depends on the size of the object, the weighted average angle is estimated. The accuracy is better than 0.02 degrees and the speed is high. The image is rotated in the opposite direction to get a skew corrected image.