Artistic Line Extraction from Indian Documents

概要

論文の詳細を見る
There are printed artistic documents where text lines of a single page may not be parallel to each other. These text lines may have different orientations or may be curved shapes. For the Optical Character Recognition (OCR) of these documents, we need to extract such lines properly. Because of multi-oriented and curved behaviour it is very difficult to extract different text lines from the document. In this paper, we propose a water reservoir principle based scheme to extract individual text lines from printed Indian artistic documents. In the proposed scheme, at first, analyzing the area of the reservoirs obtained in a component, we compute mode (portrait, landscape, reverse portrait reverse landscape) of the component. Next based on the mode and the water reservoir features like number of reservoirs, height of reservoirs, overlapping portion of two reservoirs, etc. the components are grouped into isolated or touching class. Next depending on reservoir base-area and loops of a component, some candidate envelope points are detected. Each touching component is then classified, either straight or curve type depending on the candidate envelope points of the component. Based on the type of a component two boundary points are computed from each touching component. Finally, candidate regions (neighborhoods) of the boundary points of each component are detected and analyzing these candidate regions, individual text lines are segmented.
社団法人電子情報通信学会の論文
2004-11-12

Artistic Line Extraction from Indian Documents

スポンサーリンク

概要

著者

関連論文

スポンサーリンク