@inbook{4b7686a201a54be19ec23afb15b04eb5,
title = "Word grouping in document images based on voronoi tessellation",
abstract = "Voronoi tessellation of image elements provides an intuitive and appealing definition of proximity, which has been suggested as an effective tool for the description of relations among the neighboring objects in a digital image. In this paper, a Voronoi tessellation based method is presented for word grouping in document images. The Voronoi neighborhoods are generated from the Voronoi tessellation, with the information about the relations and distances of neighboring connected components, based on which word grouping is carried out. The proposed method has been evaluated on a variety of document images. The experimental results show that it has achieved promising results with a high accuracy, and is robust to various font types, styles, sizes, skew angles, as well as different text orientations.",
author = "Yue Lu and Zhe Wang and Tan, \{Chew Lim\}",
year = "2004",
doi = "10.1007/978-3-540-28640-0\_14",
language = "英语",
isbn = "3540230602",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "147--157",
editor = "Simone Marinai and Andreas Dengel",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
address = "德国",
}