A Skew Detection Method of Mongolian Document Images
Guanglai Gao · Journal of Inner Mongolia University · 2007
In document scanning,the skew of input document images is inevitable.So the skew detection is an important step during preprocess of document analysis.A skew detection method based on least square method for Mongolian document images is proposed.In a skew Mongolian document image,every character connected component′s centroid is extracted and taken as an eigen point.Then,the least square method is applied to these eigen points for fitting straight line.Thereby,a line equation can be achieved.The line equation′s slope can properly reflect the skew angle of the document image.Furthermore,the approach also can be adapted to complex document images by connected component analysis.Experiments prove that this approach is fast with a higher accuracy.