A hybrid document segmentation method for Turkish newspapers
| dc.contributor | Graduate Program in Electrical and Electronic Engineering. | |
| dc.contributor.advisor | Sankur, Bülent. | |
| dc.contributor.author | Aktaş, M. Feridun. | |
| dc.date.accessioned | 2023-03-16T10:22:17Z | |
| dc.date.available | 2023-03-16T10:22:17Z | |
| dc.date.issued | 1998. | |
| dc.description.abstract | Today most of the information is conveyed in the form of printed papers. The range of them varies from the newspapers to formal correspondence letters, from banking documents to envelopes etc. The evaluation of document processing systems made it possible to transfer this information from the printed materials to the electronic media. To transfer and archive this information some recognition, compression and conversion techniques are used. These techniques extract the document components and process them regarding the content type. Documents are mainly composed of text and image blocks, line and drawings. This thesis is focused on the extraction of document image components for further processing. This operation is known as document analysis. Several document analysis techniques are reviewed and one of them, Recursive X - Y Cut, is modified and applied to the Turkish newspapers. This method recursively analyze the horizontal and vertical projection profile of documents and locate the most appropriate cut (horizontal or vertical) over the documents. The process recursively continues until the smallest desired blocks are found or not any appropriate cut place exists on the document. At the result, blocks that mostly contain single type of document component, are extracted. The blocks, that contains several type of document components, are fed to another segmentation algorithm.oeu(n.amtbr_i | |
| dc.format.extent | 30 cm. | |
| dc.format.pages | xiii, 96 leaves: | |
| dc.identifier.other | EE 1998 Ak7 | |
| dc.identifier.uri | https://digitalarchive.library.bogazici.edu.tr/handle/123456789/13066 | |
| dc.publisher | Thesis (M.S.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 1998. | |
| dc.subject.lcsh | Document imaging systems. | |
| dc.subject.lcsh | Hybrid computer simulation. | |
| dc.title | A hybrid document segmentation method for Turkish newspapers |
Files
Original bundle
1 - 1 of 1