Sciweavers

Share
50 search results - page 1 / 10
» The Multistage Approach to Information Extraction in Degrade...
Sort
View
ICPR
2004
IEEE
12 years 11 months ago
The Multistage Approach to Information Extraction in Degraded Document Images
Global and local adaptive thresholding techniques have been shown effective on particular types of documents. None produces consistently good results on all types of documents. In...
Graham Leedham, Yan Chen
ICDAR
2009
IEEE
12 years 4 months ago
Enhanced Text Extraction from Arabic Degraded Document Images Using EM Algorithm
This paper presents a new enhanced text extraction algorithm from degraded document images on the basis of the probabilistic models. The observed document image is considered as a...
Wafa Boussellaa, Aymen Bougacha, Abderrazak Zahour...
PAA
2006
11 years 9 months ago
Automatic name extraction from degraded document images
The problem addressed in this paper is the automatic extraction of names from a document image. Our approach relies on the combination of two complementary analyses. First, the ima...
Laurence Likforman-Sulem, Pascal Vaillant, Aliette...
CVPR
2009
IEEE
12 years 1 months ago
Robust unsupervised segmentation of degraded document images with topic models
Segmentation of document images remains a challenging vision problem. Although document images have a structured layout, capturing enough of it for segmentation can be dif´Čücult....
Timothy J. Burns, Jason J. Corso
DAS
2008
Springer
11 years 11 months ago
Accuracy Improvement and Objective Evaluation of Annotation Extraction from Printed Documents
There is an approach of annotation extraction from printed documents in which annotations are extracted by comparing the image of an annotated document and its original document i...
Tomohiro Nakai, Kazumasa Iwata, Koichi Kise
books