Document processing for adaptive page segmentation using order statistic filters
Ma, Li, 1975- · eScholarship@McGill (McGill) · 2001
Page segmentation is one of the important and basic research subjects of document analysis. Traditionally, there are two major kinds of page segmentation approaches. One is the top-down approach and the other is the bottom-up approach. Though these two approaches are been used till now, they are not effective for processing documents with high geometrical complexity and the process of splitting document needs iterative operations which is time consuming. The Modified Fractal Signature (MFS) approach which was presented in recent years can overcome the above weaknesses, however it needs to calculate modified fractal signature which makes the theory and the algorithm very complex. In this thesis, we present two new page segmentation approaches (one is the Maximum Order Statistic Filter (MaxOSF) approach, the other is the Median Order Statistic Filter (MexOSF) approach) based on the order statistic filter (OSF) which is more direct and much simpler. We use the MedOSF to remove the salt-pepper noise of the document and use the MaxOSF to do the page segmentation. In practice, they not only can adaptively process the documents with high geometrical complexity, but also save a lot of computing time.