An efficient pre-processing of mixed-content document images for OCR systems

An efficient, novel technique for segmenting document pages of mixed content into text and non-text regions is presented. The aim of the technique is to provide a pre-processing for an OCR system, so that large amounts of documents of unknown layout can be examined and the written content of such documents can be put into digital format without human intervention. The user fixes two parameters, the minimum width w of the text to be detected, and the precision ε needed (both expressed as a percentage of the image width), according to the implementation needs (default values that yield good results for widely varying kinds of documents are ε=2%, w=4ε). The method works by detecting pieces of text lines in small overlapping columns of width w-2ε, shifted with respect to each other by ε and by merging such pieces in a bottom-up fashion to form complete text lines and blocks of text lines. The algorithm is very fast and flexible and is able to work on low-resolution document pages. Experimental results are given which demonstrate the effectiveness of the method on several different kinds of documents