Sub Pixel Mapping in Degraded Document for Text Retrieval

Abstract: Segmentation of text from poorly despoiled document image is a very demanding task due to the high inters/intra deviation between the document background and the foreground text of different document scanned images. Text withdrawal from normal picture images is an emerging field in computer graphics. Extracted text contains essential information that can be used for a variety of purposes like vehicle number plate to identify the vehicle, to provide information of neighbouring to visually impair persons, preservation of information for chronological documents etc. In our project we propose new method called sub pixel mapping that addresses the issues by using adaptive image contrast and Binarization Technique. The adaptive image contrast is a mixture of the local image contrast and the local image gradient that is forbearing to text and background variation caused by dissimilar types of document degradations, binarization algorithm for each case proved to be a very difficult procedure itself. We use the case of degraded historical documents, then we apply the proposed technique called sub pixel mapping to convert the document images to the text format.

Keywords: Adaptive image contrast, document analysis, degraded document image binarization, pixel classification; sub pixel mapping.