Sciweavers

5 search results - page 1 / 1
» Enhanced Text Extraction from Arabic Degraded Document Image...
Sort
View
ICDAR
2009
IEEE
14 years 1 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...
ICDAR
2009
IEEE
13 years 4 months ago
Document Image Binarisation Using Markov Field Model
This paper presents a new approach for the binarization of seriously degraded manuscript. We introduce a new technique based on a Markov Random Field (MRF) model of the document. ...
Thibault Lelore, Frédéric Bouchara
DAS
2010
Springer
13 years 10 months ago
Handwritten Arabic text line segmentation using affinity propagation
In this paper, we present a novel graph-based method for extracting handwritten text lines in monochromatic Arabic document images. Our approach consists of two steps Coarse text ...
Jayant Kumar, Wael Abd-Almageed, Le Kang, David S....
ICDAR
2003
IEEE
14 years 6 days ago
A Model-based Line Detection Algorithm in Documents
In this paper we present a novel model based approach to detect severely broken parallel lines in noisy textual documents. It is important to detect and remove these lines so the ...
Yefeng Zheng, Huiping Li, David S. Doermann
ICDAR
2011
IEEE
12 years 6 months ago
Language-Independent Text Lines Extraction Using Seam Carving
Abstract—In this paper, we present a novel languageindependent algorithm for extracting text-lines from handwritten document images. Our algorithm is based on the seam carving ap...
Raid Saabni, Jihad El-Sana